For certain countries, however, there seem to be factors that correlate with higher levels of income and which also correlate (independently of income) with decreases in subjective well-being.  Hence, money buys happiness, but other things make you unhappy, so unhappy that they overcome the benefits of more money.  What are these things? Decreased trust levels, increased crime levels, decreased social participation, aspirations and adaptation, and many more.

The 'Happy Planet Index' ranks countries by their level of reported happiness.  What does such a ranking mean? Nothing. We do not know what cultural factors play into reporting specific levels of subjective well-being. What if Brits think it's impolite to talk about 'happiness' or 'satisfaction'? What if Germans feel that they cannot legitimately call themselves 'more satisfied than average'? What if [insert potential cultural confound here]?

What we can talk about though are trends in specific countries over time, and then compare these trends across countries.  The author of the article does both of these, she talks about how happiness was higher in the US, India and China 20 years ago, but then she goes on to discuss the rankings of particular countries and how the Latin American countries in particular rank highly.  Cultural differences? Don't mention those, they might damage the conclusion that 'development makes you unhappy'.

Some caveats: the construction of the index results from life expectancy, ecological footprint (required hectares per person), and subjective well-being.  The problem, however, is that SWB cannot and should not be included in such an index because you cannot do legitimate inter-country comparisons with SWB levels, but only changes over time. I cannot reiterate this enough and I see this error perpetrated regularly. You  cannot do inter-country comparisons of happiness levels!  Also, the flaws in the article represent flaws in the Happy Planet Index and its report. The irony is that I favour new indexes to offer as alternatives to GDP, but these indexes cannot include happiness levels for inter-country comparisons because doing so excludes too many potentially confounding factors.  Should we include, and try to explain, trends in happiness or SWB? Certainly. But it doesn't help to do so in ways that abdicate our responsibility to adequately  understand the problems of such studies.

Please excuse some of the generalizations made above about the potential proclivities of specific cultures, they were meant as illustrations and not as theories of culture. HT: My friend Marco who sent me the article from La Repubblica, I only read it occasionally otherwise.