This just in: Money does buy happiness after all

14 posts / 0 new
Last post
Stephen Gordon
This just in: Money does buy happiness after all


Stephen Gordon

I'm rather sceptical of 'happiness research', but what the hey: this study has a cool graph.

[url= recent study by Princeton's Angus Deaton (46-p pdf):[/url]


During 2006, the Gallup Organization collected World Poll data using an identical questionnaire from national samples of adults from 132 countries. This paper presents an analysis of the data on life-satisfaction (happiness) and health satisfaction and their relationships with national income, age, and life-expectancy. Average happiness is strongly related to per capita national income, with each doubling of income associated with a near one point increase in life satisfaction on a scale from 0 to 10. Unlike previous findings, the effect holds across the range of international incomes; if anything, it is slightly stronger among rich countries.

But it turns out that changes in income work in the opposite direction as levels. And the reverse is true for life expectancy:


Conditional on national income, recent economic growth makes people unhappier, improvements in life-expectancy make them happier, but life-expectancy itself has little effect.

The afore-mentioned cool graph summarising the relationship between happiness and income across countries:



I see a strong correlation at low incomes (<$10K), while the data for higher incomes that is so scattered that it could not justify choosing one model over any reasonable alternative. I.e. it seems to be consistent with existing "happiness research".

Stephen Gordon

That's not how Angus Deaton reads the data:


Like earlier studies using a smaller range of countries, I find that the citizens of richer countries are on average more satisfied with their lives than the citizens of poorer countries. Unlike the earlier studies, this effect of income is not confined to poor, unhappy countries, but extends right across the range, from Cambodia, Sierra Leone, Togo, Niger, and Chad, which share the unenviable distinction of being in the bottom ten countries both by income and by life-satisfaction, to Norway, Switzerland, Denmark, Australia, and Canada, which rank in the top ten according to both income and life-satisfaction. Each doubling of national income is associated with a near one unit increase in average life-satisfaction measured on an eleven point scale from 0 (“the worst possible life”) to 10 (“the best possible life”). If anything, the effect of national income on national happiness is somewhat stronger in the rich countries than in the poor countries.

eta: Those conclusions are based on a logarithmic transformation of income, in which going from 5k/yr to 10k/yr would be the same as going from 20k/yr to 40k/yr. If we have diminishing returns, we'd expect that curve to flatten out as income increased.

[ 01 February 2008: Message edited by: Stephen Gordon ]


What are you trying to prove with your statistics?
Do you see this as validating your own prejudices?

Polly B Polly B's picture

I don't think these are really Stephens statistics, as he points out in the postings.

Edited after reading for content....

[ 02 February 2008: Message edited by: Polly Brandybuck ]


There is surely a relationship between poverty and misery! And likewise, there's surely a relationship between life satisfaction and wealth.

I watch a lot of old movies and I notice that a lot of movies made in the depression have an obvious subtext: it is the poor who are truly happy. The rich bear burdens of which they would gladly rid themselves. I see these movies as blatant propaganda meant to counter social unrest.

rural - Francesca rural - Francesca's picture

But in terms of leaving poverty, money truly can achieve happiness.

But upon exiting poverty to "middle class" if there is still a desire to "acquire stuff" then true happiness will be alluded.

Which brings us back to the discussion on 'basic needs' - who defines what a basic need is?

I can buy $30 running shoes and they work, but if I buy $130 ones my knees are healthier, they last longer, and I can walk farther without getting sore feet.

The basic need of 'footwear' can be taken care of for $30, but the basic need of 'good' footwear takes $130.



Originally posted by rural - Francesca:
[b]But in terms of leaving poverty, money truly can achieve happiness.

But upon exiting poverty to "middle class" if there is still a desire to "acquire stuff" then true happiness will be alluded.

Which brings us back to the discussion on 'basic needs' - who defines what a basic need is?

I can buy $30 running shoes and they work, but if I buy $130 ones my knees are healthier, they last longer, and I can walk farther without getting sore feet.

The basic need of 'footwear' can be taken care of for $30, but the basic need of 'good' footwear takes $130.[/b]

I'm reacting against comfortable people moralistically and smugly saying, "money doesn't buy happiness so just go away and shut up".

But I agree about acquiring "stuff" - I think that works against happiness. I think consumerism is the road to hell!

(I also think your point about basic needs is a good one.)


Even a cursory glance at this "study" reveals it to be meaningless pseudo-scientific gibberish, there are a number of obvious methodological flaws and assumptions that demonstrate it to be unworthy of serious attention.

1) Wealth is defined as per capita GDP, this in itself is a rather meaningless measure as it says nothing of distribution of wealth. Kuwait and Norway could potentially have the same per capita GDP despite wild variations in the degree of distribution. this factor in itself raises the question of what variable exactly are the researchers claiming to be measuring it is clearly the wrong measure of central tendency to use to determine the income level of a countries citizenry.

2)Confounding variables there are many other variable that would distinquish countries labelled as "poor" from "wealthy" these are obvious historical political and social variables, levels of violence, histories or colonization and exploitation, resulting social instability and upheaval. Not only would these factors be likely to impact sense of well-being they also likely to be related causally to income levels and productivity and wealth distribution.

3) Cultural assumptions-when discussing a culturaly bound construct such as well-being it is unlikely that western definitions of well-being are likely to be the same as other cultures. People in western countries are have enough difficulty determining what constitutes well-being, even prior to making claims of universality.

4) Problems with self-report data, there are a number of problems determining the accuracy of self-report data one that would be particularly relevant for this study would be social acceptability. In North American culture we are taught to judge our very value as decent and acceptable human beings by the amount that we earn or the work we do, money=success=happiness, to admit to be unhappy or dissatisfies in this culture is to confess that you are being a failure with all the attendant shame. It is highly unlikely that people in this culture are going to be acknowledging their failure to complete strangers.

5) Measure of well-being- the measures of well-being seem rather simplistica and clustered for the most part around self-reported assessments of physical health and access to health care. Are these true measures of global well-being across cultures I doubt it. The researchers make a bizarre statement about this


While these measures are far short of those that might appear in a comprehensive health and economics survey for a single country or a group of similar countries, they have the great advantage of having been asked in exactly the same way in all of the countries.

Basically they're saying that our measures may be invalid but at least they are consistently invalid?

They actually follow this up with an even more bizarre conclusion


The question is not, therefore, whether these self-reports are adequate measures of individual health—they clearly are not—but whether they provide useful measures of population health

So somehow if you aggregate inaccurate measures of individual health they magically become true for a population, that looks like the sort of logic one would expect from economics.

There a number of other problems that crossed my mind but I think that will do for now.

So I think it's safe to conclude that this study isn't even looking at accurate measures of wealth or well-being so I'm not really sure what the point is.

Stephen Gordon

As I said, I'm pretty skeptical of the 'happiness' research agenda: self-reported measures of happiness are much less informative than observations about how people actually behave.

But it's important to give the 'poor=happy' point of view a whack upside the head from time to time. If there were no connection between poverty and happiness, then poverty wouldn't be seen as a problem.

The Wizard of S...

Money may not buy happiness, but it will buy Guitar Hero III: Legends of Rock. Right now I'm dueling the Devil, who's name is Lou for some reason, playing "The Devil Went Down To Georgia." It's hard as hell, but pretty damn fun.

Boom Boom Boom Boom's picture

Half a million bucks would make me happy - I could rebuild this place with an indoor heated pool, sauna, equip the house and garage with a couple of large generators for when the hydro goes out, purchase a small tractor and snowblower to move all the @#$%!!! white stuff to the edge of the property, and so on. You get my drift. [img]biggrin.gif" border="0[/img]


Happiness comes in surviving IN SPITE of the system.
In not buckling to it. In not playing to ITS rules.
Having money, I've found, really DOES help, but much can be done without nurturing parasitism and its social costs.