Understanding "The Pie Chart" in The How of Happiness (Ed Batista)

The-How-of-Happiness

Sonja Lyubomirsky, a professor of psychology at UC Riverside, is one of the leading researchers on happiness, and her books The How of Happiness [1] and The Myths of Happiness [2] offer a wealth of valuable insights on the subject. (Both are on my syllabus at Stanford for The Art of Self-Coaching. [3]) One of the most significant topics Lyubomirsky addresses is the extent to which different factors contribute to happiness, a major focus of positive psychology in recent decades. But it can be easy to misunderstand Lyubomirsky’s discussion of this topic in her earlier book, and it’s important to clear up any misconceptions in order to make effective use of her research.

The widely cited "pie chart" from The How of Happiness (referenced on the book’s cover above and shown at right) should correctly be understood as representing the amount of variance in happiness among individuals in Lyubomirsky's research that can be predicted by heritable traits, life circumstances, and intentional activities.

It's easy to get the mistaken impression that this chart represents the relative amounts of happiness that are derived from these sources for individuals in the population at large. But this research does not (and cannot) tell us how much of a given individual's happiness is derived from each of these sources.

Ed Diener, a professor of psychology at the University of Utah and the University of Virginia, as well as an emeritus professor at the University of Illinois, has addressed this potential confusion:

One misunderstanding of percent of variance figures applied to the causes of human happiness is that the numbers are presented as though they apply to individuals. In other words, 45% [of] an individual's happiness is said to come from inborn temperament, 12% from demographics, and so forth. The percent figures are derived, however, from the amount of variance between individuals that is "explained" or predicted by specific variables. These numbers tell us little about the absolute importance of the variables, because people in a sample may or may not differ to the same degree on each of them, and they do not tell us how individuals would change in well-being if they were to change on these variables. It is inappropriate to interpret the figures as applying to how much an individual's subjective well-being is derived from the various causes because the numbers are derived from variation between people in specific samples. [emphasis original] The percentages might be interpreted to mean that if one were to improve one's demographics from terrible to great, one's happiness might increase by 12%, but this is a misunderstanding of what the figures mean. It is important to recognize that the percent of variance between individuals, due to various causes of happiness, depends on the range and variation between people on this factor and has no necessary connection to what might be important to altering a person's happiness...

Happiness is sometimes said to be about one-half heritable, but this statement can be easily misunderstood. It is a descriptive statistic based on particular samples in particular life circumstances and might not apply to other samples--for example, ones in which life circumstances are more variable across people. That is, the "heritability" of happiness is not constant across samples. This point surprises some people because they do not realize that heritability is not the same as genetic effects.

Pie-chart numbers sometimes lead researchers to view some variables as more important than other variables, and this is often mistaken. Furthermore, these figures are sometimes offered to the public as a guide to what might be the most worthwhile to change in order to achieve greater happiness. However, the causes for change in an individual's happiness might diverge from what causes differences in happiness between individuals. For instance, one person might gain an enormous boost in happiness from becoming religious, even if the amount of individual differences in happiness due to religion in a population is modest. The pie-chart way of thinking is seductive, because it is clear and simple, but it can easily lead us to think about the causes of subjective well-being in misguided ways...

The percent of variance figures derived from sample statistics do not apply to individuals. [4]

Another psychologist who addressed this issue was the late Christopher Peterson of the University of Michigan:

One of the frequently cited conclusions from positive psychology research is that happiness results from a combination of genetics, circumstances, and voluntary activities. This is reasonable enough. Indeed, it is a virtual tautology that applies to most any human characteristic.

Some positive psychologists go further and propose a happiness formula, typically a weighted sum of its components, with weights based on research with large samples of individuals. A representative set of weights is 50% genetics, 10% circumstances, and 40% voluntary activities. Again, this is reasonable enough, reflecting the research literature as I read it, although the exact weights are always a function of the samples from which they are derived.

So where am I going? To the conclusion that it is thoroughly unreasonable to think that we can parse the happiness of an individual, in the moment or in general, in the same way that we can parse the happiness of samples of individuals... [emphasis original]

It is not clear to me whether positive psychology authors who present such formulas intend these formulas and their weights to apply to individual people or to the specific moments of happiness that individuals experience. I do know that their readers often make these leaps because I encounter this notion with incredible frequency among my students who have read popular trade books on happiness. I spend a lot of time trying to explain heritability to them. [5]

To the point raised by both Diener and Peterson regarding heritability and genetic effects, genetics scholar and journalist Razib Khan provides a helpful explanation:

When someone tells you that height is 80% heritable, does that mean:

a) 80% of the reason you are the height you are is due to genes
b) 80% of the variation within the population on the trait of height is due to variation of the genes

The answer is of course b. Unfortunately in the 5 years I’ve been blogging the conception of heritability has been rather difficult to get across, and I regularly have to browbeat readers who conflate the term with a. That is, they assume that if I say that a trait is mostly heritable I mean that its development is mostly a function of genes. In reality not only is that false, it’s incoherent. Heritability is addressing the population level correlation between phenotypic variation and genotypic variation. In other words, how well can genetic variation work as a proxy for phenotypic variation? What proportion of the phenotypic variation can be accounted for by genotypic variation? [6]

Note that in the text of The How of Happiness Lyubomirsky is careful to specify that the research she's discussing predicts the variance in sources of happiness among individuals, not the relative weight of those sources for a given individual. That said, the illustration of the "pie chart" with the label "What Determines Happiness?" can contribute to the misunderstanding that Diener and Peterson warn against.

Some comments in the book can also be misinterpreted--Lyubomirsky writes, "As we can see from the pie chart, changes in our circumstances, no matter how positive and stunning, actually have little bearing on our well-being." [7] This is accurate with regard to the variance in the sample, but it could be read as advice to an individual, in which case it may well be inaccurate for any given person. To be clear, I’m not blaming Lyubomirsky for misreadings of her work, but it's essential to have a firm grasp on what her research tells us and what it does not.

With that caveat, here's a summary of my current understanding of the research on happiness:

Happiness matters. Relative to unhappy people, happy people do better on any number of measures, from income to longevity.
More happiness isn't necessarily better. (As Peterson notes, while the happiest people are more successful at close relationships, people who are slightly less happy--although not unhappy--are more successful at work and in school. [8])
We've traditionally understood happiness as a function of life circumstances, such as income. This view overlooks two important sources of happiness: genetics and intentional activities.
Our life circumstances--including income--may matter less than we think. Life circumstances predict relatively little of the difference in happiness between individuals in much of the research, although the role life circumstances play in the happiness of any given individual will vary substantially.
There is a correlation between income and life satisfaction, but this does not imply causation. There is also a point of diminishing returns, although the precise inflection point varies from study to study. (As Diener notes, the relationship between money and happiness is complex and still not fully understood. [9])
Hedonic adaptation diminishes the impact of life circumstances on our happiness over time, although changes in our circumstances may alter our baseline level of happiness. The specific impact of any changes and of our ability to adapt to them will vary for any given individual, but it's worth noting that our traditional narrative about the sources of happiness and their longevity are often incorrect. [10,11]
Our genes do play a role in our happiness, but even our "set point" is not fixed. [12]
Intentional activities are an often overlooked source of happiness. These are the "happiness strategies" discussed by Lyubomirsky, and they predict a significant amount of the difference in happiness between individuals in much of the research, although, again, this will vary for any given individual, and subsequent research has shown that the overall impact of intentional activities is likely smaller than originally estimated. [13]
Although our genetic set point isn't permanently fixed, due to the impact of our environment and developmental experiences on which genes are expressed and how, our genes are, obviously, predetermined. And although life circumstances do have an impact on our happiness, 1) that impact may be substantially less (and shorter-lasting) than we imagine it to be, and 2) changes in life circumstances are often difficult and occasionally impossible to bring about through deliberate effort. In contrast, small-scale, intentional activities may have a much larger impact on our happiness than we imagine, and by definition they are easy to put into practice.
Consequently, many people may find that engaging in some of these intentional activities have a surprisingly large (and persistent) impact on their happiness. We all respond differently to various activities, so it's important to choose the right ones, and here Lyubomirsky's Person-Activity Fit Diagnostic is a useful tool.