Economists (don’t) prove that money can buy you happiness… And news outlets prove they’re crap.

This post was chosen as an Editor's Selection for ResearchBlogging.orgLast week, Science published the results of a study [pdf] that positively answers once and for all the question that every philosopher (and non-philosopher) since Plato has asked themselves: Money can buy you happiness!

However, the results might not be as clear-cut as they seem. The article and its media coverage raise two separate points: 1) The popular media manages to report on science stories in a way that completely misses everything that makes the story worthwhile. 2) The study itself has an interesting design flaw which renders the conclusion far less surprising than the authors think.

News outlets all over the world rewrote the press-release mostly focusing on a minor part of the article which ranks the 50 states of America from happiest to least happy. This ranking is a simple analysis of publicly available data which is an all but irrelevant aspect of the research. The authors merely use this data as a tool to show that money can buy you happiness. If this much-reported-on aspect of the research was the main point, I doubt it would have got published anywhere, let alone in Science.

While we would dearly love for scientists to be able to discover the secret to happiness, psychological research has shown us that it is much more difficult than it may seem to figure out what makes us happy. In fact this study – and others of a similar ilk – actually only manage to tell us, at most, the secret to happiness in one particular part of our life.

Happiness studies: an emerging discipline

Happiness studies have exploded in the last few decades. It is emerging as a new and exciting interdisciplinary field encompassing psychology, philosophy, economics and neuroscience, complete with its own journals and research centres and it attracts record numbers of students to its courses all over the world.

At its very core, there are two basic questions about happiness that the field seeks to answer: What is it? And how do you get it?

Economists Andrew Oswald and Stephen Wu from the University of Warwick in the UK have attempted to find out what makes us happy. Or, more accurately, they have attempted to find out what makes us feel a sense of satisfaction with our lives as a whole.

“Life-satisfaction” is a common definition of happiness among both philosophers and psychologists. It is measured by asking the subject, “All things considered, how satisfied are you with your life as a whole?”

Many philosophers defend this definition of happiness by arguing that other definitions which usually equate happiness with pleasure are too superficial to count as a definition of happiness. They say that life-satisfaction captures what a lot of people mean when they talk about “leading a happy life”.

So, does money buy you happiness?

So how do you lead a happy life? Can you buy life-satisfaction with money?

Oswald and Wu used the data from a study that asks 350,000 people each year in the USA over 100 questions about their life conditions. The study collects data from all over the USA and began doing so 25 years ago. The survey provided the researchers with an amazing 1.3 million responses to work with.

From this data, the two economists gave each state in the USA a score for how happy its residents were each year.

Oswald and Wu were then able to compare that data with a measurement of what is called the “quality of life” of the residents in each state. Quality of life measurements assume that “people enjoy having high income, and spatial amenities such as sunlight hours, but they dislike disamenities such as traffic congestion.”

Quality of life measurements assume that some number of dollars in a worker’s paypacket can offset the unpleasantness of things like poor air quality or few hours of sunlight.

Astonishingly, they found a very high correlation between life-satisfaction and quality of life. The higher a state’s quality of life, the higher its residents’ life-satisfaction.

If accurate, the conclusion you can draw from this study is that money can buy you happiness! By its definition here, “quality of life” can be bought: If you have a big enough pay packet, you will have a high “quality of life” because the pay will offset any negative contributions from things life traffic congestion. Now, if quality of life is equivalent to life satisfaction (as is suggested by the very high correlation), then you must be able to buy life-satisfaction too!

Problems and design flaws

One of the main arguments philosophers present against defining happiness as life-satisfaction is the discovery by psychologists that judgements of life-satisfaction are incredibly dependent on irrelevant factors such as which questions were asked prior to the life-satisfaction one.

In one study, psychologists found that if you ask subjects how happy they are with their lives and then how happy they are with their marriage, the responses to the two questions were only correlated about one third of the time.

However, if you ask them first how happy they are with their marriage and then how happy they are with their lives, the responses are correlated roughly two thirds of the time.

The idea is that respondents are not able to consider their entire life when judging their satisfaction and so focus on some manageable aspect of it. Which aspect they focus on can be controlled by first asking questions about that aspect of their life.

As one review of the literature on context effects concludes, “these influences may be so pronounced that researchers may draw opposite conclusions about the same substantive relationship, depending on the order in which they ask the relevant questions.”

In the case described above, this was exactly what happened. If respondents were asked how happy they were after being asked about their marriage, they mostly focussed on how happy they were with their marriage.

So the big question is, what kind of life-satisfaction data did Oswald and Wu use?  Were those data subject to identifiable context effects?

In short, yes. What Oswald and Wu don’t mention is that prior to the life-satisfaction question there are over 100 questions asked, almost all of which relate to the respondent’s  health. You can see the questionnaire for yourself here.

The context in which the life-satisfaction question was asked was so skewed that the results are best interpreted as responses not to life-satisfaction, but to a question more like “All things considered, how satisfied are you with your health?”

So can money buy happiness? Well, I don’t think Oswald and Wu manage to show that. What they perhaps do show is that in the USA, money can buy you satisfaction with your health. But that is hardly as surprising as money buying you happiness.

Oswald, A., & Wu, S. (2009). Objective Confirmation of Subjective Measures of Human Well-Being: Evidence from the U.S.A. Science DOI: 10.1126/science.1180606

Schwarz, N., Strack, F., & Mai, H. (1991). Assimilation and Contrast Effects in Part-Whole Question Sequences: A Conversational Logic Analysis Public Opinion Quarterly, 55 (1) DOI: 10.1086/269239

Schwarz, Norbert, & Strack, Fritz (1991). Context Effects in Attitude Surveys: Applying Cognitive Theory to Social Research European Review of Social Psychology, 2