On Happiness

melted_snowball was writing about Daniel Gilbert's interview on Tapestry about happiness.

Gilbert says: paraplegics are just as happy as lottery winners. People raising children are less happy than people who do not have children. These may be true in some sense, and Gilbert does have interesting things to say. melted_snowball disagrees with his premise, that by comparing people against each other, you can find a meaningful "average happiness" that is useful for measuring quality of life.

I'd like to disagree with something else: his slippery definition of happiness.

I've not read his book yet, and I hope to as soon as the public library tells me a copy is free; and I'm willing to change my opinion after I've seen the book. But lilibet pointed me toward his TED talk that suggests people manufacture happiness- they tend toward a baseline "happiness"; they imagine the past as if they were closer to their current level of happiness; they don't predict their future happiness at all well (such as whether more income will make them happier.) What is this "happiness" as Gilbert defines it in these two clips?

Getting what you want.

Contrasted with, say, being foiled in what you want by an experimenter. At least as far as I can tell. That's not "happiness," that's... satisfaction? Lack of dissonance? If that's the definition he's actually using, I don't think it's useful at all. Optimizing for getting what you want won't make the world happier, it'll make it spoiled.

I'm likely over-reacting to Gilbert's pop-science presentation of his argument; if so, I hope his (er, pop-sci) book will make it somewhat clearer what he's measuring. And tomorrow at work I'll take a look at his paper, Affective Forecasting: Knowing What to Want, which is stuck behind a journal's firewall. The joy of University library access!)

While looking around online, I found the work of John Helliwell, an Economist at the University of British Columbia, and his definitions seem quite a bit more nuanced; they're based around happiness and well being.


A basic assumption in economics is that people want to maximize their utility, or well-being, and economists have long assumed per-capita income and wealth to be reasonable measures of this. However, recent research in psychology shows many additional factors boost people's sense of well-being as much as, if not more than, their monetary worth.

Among these factors is what many refer to as social capital, or "the networks and norms that facilitate collaborative action," according to Dr. Helliwell. These include civic engagement — participation in community organizations, for example — and social interactions like those with friends and family. Other factors linked to well-being are trust (in society in general and in specific domains like the police, government, neighbours and co-workers), employment (whether paid or not), good health, a stable family and effective, high-quality government.

Income does have an effect on well-being up to a certain point, but this effect diminishes at higher income levels. What matters more is relative income — people are less happy when they think that those around them have a higher income than they do. Age, too, affects well-being, with both younger and older people happier than those in their middle years (40 to 50 years old). Dr. Helliwell is not sure what accounts for this, but hypothesizes it may be related to issues of work-life balance.

Of interest to academics is that education doesn't seem to affect well-being directly. Dr. Helliwell hastens to add, however, that it does affect well-being indirectly through factors such as income, health and civic engagement — variables that are all known to be correlated with education.


I think I will work my way through Helliwell's paper on well-being and social capital while I'm waiting for Gilbert's book to show up at the library. As far as I've read, it seems quite a bit more satisfyingly rigorous.

[Edit to add: Many of Gilbert's papers are available on his website, I just didn't read the pale-gray text at the top which said to click on the orange bullet-point to download each paper. Heh.

Anyhow, both of the articles I've just read (the one I noted above about Affective Forecasting, and one called "How Happy Was I?") used self-reporting of happiness on a numerical scale. I wonder whether I'm just biased, or is this discovery making me unhappy?..]