People are silly

I've been reading a lot of papers in the domain of "happiness economics", which is mostly concerned with the so-called happiness paradox: while all measures of wealth, income and standard of living have increased considerably over the past 50 years, we haven't seen any increases in people's self-reported happiness. Economics has given this problem a bit more attention in recent years, but probably not as much as it deserves. After all, we're supposed to be concerned with the allocation of scarce resources in the best possible manner, i.e. to maximize people's utility or happiness. If the link between income and happiness is so tenuous, the optimal allocation of resources doesn't seem very important. And if economic growth and prosperity don't make people any happier, then there goes a large portion of the profession that can be dismissed as "mostly useless".

After reading approximately a million of these papers you start to see a pattern. This is pretty much how they go:

FACT 1: People aren't any happier now, with all their cell phones and interwebs and enormous houses and whatnot, than they were 50 years ago, or whenever the subjective well-being survey was first administered.

Most of the empirical studies try to measure what they refer to as "subjective well-being", which is pretty much just how happy you think you are. It's proxied by questions like "Would you say that you are generally very happy with your life, happy, neither happy nor unhappy, unhappy with your life, or very unhappy." The responses are aggregated, and the happiness of the population is measured by the proportion of people who answer in each of the categories. And all the within-country, across-time studies find that these proportions remain constant.

FACT 2: Income only really matters at levels of subsistence or extreme poverty.

What they do find is that for countries where average income is less than $12 000 a year, economic growth in the form of higher income over time does lead to higher levels of reported happiness. This is generally unsurprising.

FACT 3: Within-country variation in income is correlated with happiness.

That is, the rich people in the country are happier than the poor people. Also unsurprising.

Traditional economics doesn't like facts 1 and 3 taken together, because they seem to contradict one another. If rich people are happier in the population, then making people wealthier over time should make them happier. Enter happiness economics, which introduces the notion of a positional good. The papers that talk about positional goods usually cite surveys that ask people some variation on these questions:

1. Choose between world A in which you have $50,000 in real purchasing power, while everyone else has $25,000, or world B in which you have $100,000 and everyone else has $200,000.

2. Choose between world C in which you have 2 weeks vacation time and everyone else has 1 week, or world D in which you have 4 weeks vacation and everyone else has 8.

Approximately half of people will prefer world A to world B, while about 80% choose D over C. One of the first economics books I read used this example, and I was a little bit taken aback about the responses to question 1. I suppose it's understandable how some people might value being high up in the income distribution so much as to forego half of their salary, but more than 50%? I found that a little unsettling. When you think about it, going from world A to world B is like saying "How would you like it if I doubled your income and quadrupled the income of everyone else?". I'm pretty sure I would take this offer without hesitation. Fortunately the same result doesn't occur with the second question, because leisure is a less conspicuous form of consumption. Most people can't easily demonstrate to their neighbours how much vacation time they have, and it's less susceptible to envy. This is pretty much the opposite of how I feel about things, having chosen the low-consumption, high-leisure trade-off of grad school.

Economics doesn't like this because it's too context-dependent. In almost every economic model, individuals only care about their consumption bundle. $100,000 consumed in a world where everyone is poorer than you should be just as good as $100,000 in world where everyone is richer. But with positional goods, your position in the distribution matters (to some extent) more than the absolute value of the good. What's bad about this is that it can lead to arms races, in which everyone is trying to move up the distribution and ultimately no one is better off. Papers often quote the anecdote of the job candidate with a really expensive suit that leaves a favourable impression on the employer. This prompts every candidate to purchase expensive suits, destroying the benefit it conferred in the first place (as a form of distinction). I think this is the way a lot of sports environments work too. A new swimsuit comes out which is supposed to be super fast, but if all your competitors have it too, you're not really that much better off.


One of the papers asked a bunch of these hypothetical questions. The numbers here represent the percentage of people (sample size 146) who would accept worse values of these things in absolute terms if it meant having a better value than everyone else. I can understand the attractiveness measures being high on this list, but the ranking of "Child's intelligence" is a little disturbing to me. Not that I'd want my kids to be dumber than everyone else, but if I had to pick between having stupid kids in a community of morons or having somewhat intelligent offspring among a group of brilliant children, I would have to opt for the latter. The "Papers to write" is also a little baffling. Who are these 43.8 people who would rather write a greater number of papers if it means their colleagues are writing more than them? And finally I have to take issue with the "Own education" outcome. Depending on the degree and subject, education is the quintessential worthless signaling tool. For most people the only value that a university education provides them is as a signal to employers that they are reasonably intelligent people. Beyond the more or less practical disciplines like engineering, nursing, etc. or for those who want to go on and teach, how many people actually make use of the things they learned in school? Seriously, think of all those english and history degrees whose sole reason for existence was to label someone as "university educated". Maybe I'm biased by my belief that university is mostly a sham, but I think "Own education" should top that list. It makes no sense not to prefer a world where my B.Sc. makes me the most educated person, to a world in which I have my Ph.D. and an assortment of postdocs under my belt, but so does everyone else. People are silly.