When it comes to measuring happiness, it is a lot cheaper and easier to ask high-level evaluative questions than it is to ask about specific feelings and activities. While I have voiced my concerns about these questions, happiness snapshots are better than no record at all of our happiness and we have more information on them than for any other measure. As a result, some of the evidence I cite in later chapters will refer to life satisfaction and so it is worth briefly considering some of the evidence.
Probably the best international data on life satisfaction come from two studies carried out in the UK and Germany. In each data set, for about the past 20 years, the same 10,000 or so people have been asked about their life satisfaction alongside lots of other questions about themselves and their lives. These are called longitudinal data because we have multiple observations on the same people over time. Economists like me generally prefer longitudinal data because they allow us to see how each individual’s happiness changes in response to good or bad life events. A few years ago, I led a comprehensive review of this literature and our conclusions were that life satisfaction ratings are higher for those who:
a. are wealthier (especially when compared to people who are like them)
b. are young or old (being in your forties and fifties is a bad time for life satisfaction)
c. are healthier
d. have lots of social contact
e. are married (or at least cohabiting)
f. are a little more educated (having a degree is good but you probably shouldn’t get a PhD if you want to maximise your life satisfaction)
g. are religious (it doesn’t matter which religion)
h. are in work
i. commute a short distance to work
Since our review, some further details have been added to some of these effects. Money appears to matter a lot when you are poor, but the impact on life satisfaction of each additional dollar shrinks — though never to zero, as it appears to do for daily mood. We need to be careful here, though, because income does not only directly affect life satisfaction; it also indirectly affects happiness through its impact upon other inputs that affect life satisfaction. Richer people are generally more likely to have more friends, get married, be in better health, and so on, all of which improve life satisfaction. So rather than isolating the effect of income, which economists tend to do, we need to sprinkle its effects across all the other inputs into life satisfaction. When this sprinkling takes place, the effect of income on life satisfaction is much greater than found previously in the literature because we are picking up its indirect effects as well as the direct effects that come from having a bigger bank balance.
It has been suggested that the U-shaped relationship between life satisfaction and age might be because of expectations: as young people get older, they expect to be more satisfied with their lives than turns out to be the case, but once they get through their fifties, they expect to be less satisfied than they end up. Having children delays the onset of the downward move on the U by two decades but this is due to differences in income and education among people with and without children rather than the children per se. It would also appear to be the case that life satisfaction takes another dip again once you are lucky enough to reach 75. There is some evidence that those who say they are at the top point on a life satisfaction scale, such as “10 out of 10,” are likely to be older (as well as poorer, less healthy, and less educated) than those who say they are “9 out of 10.” Such findings can lead us to further question just what the term satisfaction is getting at.
Our ratings of life satisfaction are also affected by “internal” attributes, like personality and genes. Sociable people (high in extroversion) tend to be the most satisfied with their lives, and anxious people (high in neuroticism) tend to be the least satisfied. It is important to keep in mind, though, that personality is not entirely fixed and can change over time. The effect of genes, in particular, has led some to believe that we each have a set point of happiness that we fluctuate around but always return to. But this is not supported by evidence, because some events, like unemployment and disability, can permanently lower satisfaction with life. And for some people, marriage can have long-lasting positive effects. In the next chapter, I’ll consider in more detail the evidence on what we get used to and what we do not.
There are also some exciting new data that, in time, will enable us to make more confident claims about the associations between different measures. The Office for National Statistics (ONS) in the UK, which gathers a range of data about economic growth and about how well life is going in other ways, is now trying to monitor national happiness in a number of ways. The ONS surveys are now asking nearly 200,000 people per year across the UK about their happiness, using four main questions:
1. Overall, how satisfied are you with your life nowadays?
2. Overall, how worthwhile are the things that you do in your life?
3. Overall, how happy did you feel yesterday?
4. Overall, how anxious did you feel yesterday?
All responses are on a 0 to 10 scale where 0 represents “not at all” and 10 means “completely.”
There are similarities in how age affects responses to the four questions: there is a general “misery of middle age” confirmed in all cases. Those aged 45 to 49 report the lowest life satisfaction, worthwhile activities, and happiness and those aged 50 to 54 report the highest anxiety. There are also some interesting differences in the responses across different groups. Women are happier on all three positive measures but they also report more anxiety. These findings are broadly in line with other research looking at gender differences, although the gap between the life satisfaction ratings of men and women appears to have been narrowing over the last few decades.
I would like to raise a word of caution in comparing the results across different studies – and sometimes even within studies. When the first ONS data were released in 2012, the questions were asked either during face-to-face interviews or over the telephone, and so we looked to see whether the mode of administration made a difference to reports of happiness. There was a difference – but in which direction? If you are anything like me, you would think that there is an inclination to appear happy when someone is sitting opposite you and so the participants interviewed face-to-face would report being happier than their telephone survey counterparts. You know what’s coming – we found the exact opposite: the telephone folk were happier.
We don’t have a robust scientific interpretation of why this occurs, but when I discussed our results with Danny Kahneman, he suggested that people can’t lie to your face whereas they can inflate how happy they really are on the phone. More research is needed but in addition to knowing what questions people have been asked in happiness surveys, we need to know how they have been asked.
Happiness by Design is available on Amazon UK and Amazon US.