This week David Cameron has announced his plans to develop a national happiness index in an attempt to supplement purportedly insufficient present measures of national well-being, such as GDP. Cameron is not the first leader to make this move (Sarkozy and Obama beat him to it) and he will not be the last. Indeed, for as long as governments find themselves having to navigate through troubled economic waters there will be enough motivation to steer the public’s attention towards potentially happier shores.
Cynicism aside, many argue that this general movement towards alternative measures of welfare has the potential to revolutionise evaluating the well-being of individuals, communities, and nations, and improving policy accordingly (an optimism shared by Lord Layard’s Movement for Happiness). So it is worth taking seriously. The remainder of this piece considers how subjective well-being research (often referred to as happiness research) needs to change if it is to have such an impact.
Many articles written in response to Cameron’s proposed happiness index simply express bemusement over the idea of being able to measure people’s happiness, especially being able to do so on a national scale. (A couple are here and here.) But, subjective well-being researchers have been answering these kinds of questions and improving their methods for studying happiness for over 50 years. By now, we are seeing world-wide collaboration in the field of happiness studies on a scale never before undertaken (featuring psychologists, sociologists, economists, and biological scientists, as well as health and business practitioners) providing measurements comparing individuals, communities, and nations on a vast variety of scores.
A frequently used method is the “Satisfaction with Life Scale,” which asks individuals questions along the lines of “How satisfied are you with your life-as-a-whole?” The finding from these methods can then be compared to what family members, friends, and colleagues of the respondents have to say about the levels of happiness, and to surveys of the same subjects over longer periods of time. Of course, people’s self-reported life satisfaction can seem widely off the mark (the chronically ill may report to be as satisfied as those who are perfectly healthy), but researchers are not searching for how satisfied people should be with their lives, but rather how satisfied people feel with their lives.
With this, I have no problem. However, I think that the primary focus on the notion (and measurement) of life satisfaction is a major problem for subjective well-being research. Now, there are other methods of measuring subjective well-being. Nobel Prize-winning psychologist Daniel Kahneman, for instance, has developed both the “Experience Sampling Method” and “Day Reconstruction Method,” which attempt to focus subjects on their present emotional experience (which are then aggregated over time) rather than provide global evaluations of life satisfaction. Yet, because they are relatively inexpensive and easy to use, life satisfaction survey questions remain the most frequently used method of measuring subjective well-being, and will without doubt be the focus of Cameron’s happiness index.
So what’s the problem with researchers primarily focussing on life satisfaction? Philosopher Dan Haybron has plenty of reasonably things to say on this matter and I will briefly outline a few of his points here. In general, the problem with life satisfaction attitudes is that they do not seem, by their very nature, to be the sort of thing we should expect to track well-being in any straightforward way. Haybron convincingly argues that life satisfaction attitudes can quite reasonably depend on factors that even individuals themselves would consider irrelevant to their well-being: whether I am satisfied with my life depends partly on how well I see my life going relative to what I care about, but it also depends heavily on lots of other things that have no bearing on that question.
This is because individuals’ own sense of how things are going in their lives will normally leave the question of whether to be satisfied with their lives highly underdetermined, for two reasons. First, people’s values are often incommensurable and impossible to aggregate in any principled way. Our lives are complicated, and summing up their overall quality in a single number is bound to involve a great deal of arbitrariness. Second, life satisfaction is not just a well-being judgment: it is a judgment about whether one’s life is good enough – satisfactory. So it embodies at least two judgments: how far you’re getting what you care about, and where to set the “good enough” point (do you need to get 80 percent of what you seek to be satisfied, 60 percent, or 20?). It is doubtful that most people have any firm sense of where that point should be, and still more doubtful that they should. (If I resolve not to be satisfied unless I get at least 72 percent of the items on my life list, would that be reasonable, and more to the point: should anyone care whether I cross that line?)
As a result, all manner of factors other than well-being can quite reasonably drive our life satisfaction judgments: ethical norms (e.g., valuing gratitude, or non-complacency), perspective shifts (as might happen at a funeral or a school reunion), pragmatic factors (e.g., self-motivation, cheering oneself up), positivity biases, etc. Since my sense of how things are going, and my convictions (or lack thereof) about how good is good enough, may leave it wide open whether to be satisfied or not, I can move the bar to suit my purposes. Haybron provides the following example: Prior to getting cancer, say, I might have set it pretty high, emphasizing my values of non-complacency and striving, and not been satisfied despite getting most of what I want out of life. After the diagnosis, I decide to emphasize gratitude and fortitude, and keeping a positive outlook, so I low the bar considerably. Thus I wind up more satisfied with my life after getting cancer, even though I am quite certain my life is going worse for me. My life satisfaction, in this case, contradicts my own sense of my well-being. Yet simply by asking about my life satisfaction, subjective well-being researchers are missing out on this fact. The (I believe, small) extent to which life satisfaction contributes towards subjective well-being, researchers have things covered, but to the (I believe, large) extent that it does not, the study of happiness will be left wanting.
Thus, before drawing general conclusions about the happiness of Britain, subjective well-being researchers will need to put a great deal of effort into combining different methods of measurements, including self-reports, peer reports, observational methods, longitudinal studies, experimental studies, and physiological and other methods so as to refine the interpretation of responses to surveys. Once this has been done sufficiently, then we might have a (scientific) revolution on our hands.