Monday, November 12, 2007

Income and Happiness: A Theory

To my knowledge, there is very little evidence that shows that happiness increases with income. The evidence tends be the following:
  1. People in rich countries are not happier than people in poor countries.
  2. There are lots of stories of people who have gotten rich and remained/become very troubled.
  3. Within countries, there is a little-to-no relationship between income and happiness.
  4. Economic growth doesn't make people happier in a given country.
Many take this as evidence that there is little to no relationship between income and happiness. Yet this is completely counterintuitive since everyone thinks they'd be better off if they had more money. Daniel Gilbert, in Stumbling on Happiness, explains this inconsistency by arguing that humans are terrible at making predictions. I think Gilbert's story is a subplot to the main story.

As an undergrad, a psychology professor told my class about an experiment in which someone in a bad mood was inserted into a room full of other people. The room was infected very soon afterwards. Even those who didn't interact with this person were adversely affected. The punchline? There are happiness externalities. Further, the negative happiness externalities are more powerful than the positive ones. It's easier to destroy happiness than to create it.

Also, think about the times you've tried to cheer up a friend who was feeling down. More often than not, I'd bet the night ends with you both in a bad mood. Also, think about the last time you walked through a rough area of a city. I bet, like most folks, you kept a solemn expression on your face. Think that didn't affect your mood?

This idea can explain why getting rich doesn't make people that much better off. They still have just as many poor (independently unhappy) folks around bringing their mood down.

This also explains why rich countries aren't happier than poor countries. Or why economic growth doesn't bring happiness. With economic growth comes greater inequality. That is, rich countries still have lots of poor (independently unhappy) people.

Is this idea consistent with the fact that relative income matters? I think so. Just because I want to have a rich happy neighbor doesn't mean that I don't want to be even richer than he is.


Related Links:
Update: I didn't realize it at the time, but this post is totally influenced by Malcolm Gladwell's The Tipping Point which I had just barely started reading the night before. Now that I'm further in the book, the influence is obvious.

6 comments:

Anonymous said...

Relative income does matter as far as "relative" relative income, and does affect the mood of other relatives. Can be very depressing! Leaving the ,"scene of the crime", may be necessary for inner peace! Out of sight, out of mind!

Will said...

Aren't there serious methodological issues with those happiness studies. Do you really know people are not happy by just asking them? and isn't there an issue with using bounded scales?

Maybe happiness is increasing but when people answer surveys they use the relative happiness around them as an anchor. Suppose you're asked to rank your happiness on a 1-10 scale. How do you know what an 8 or 9 on that scale means? Maybe you look around you, figure average happiness is about an 8, you don't feel that much happier than average so you give yourself an 8. If everyone did this, then even in an environment of increasing happiness the happiness measures wouldn't change.

Also, the scales are bounded above. Incomes have quadrupled in the last 50 years, but its impossible for the happiness scores to do so. In the extreme case, happiness would get stuck at 10 and the headlines would still read "economic growth doesn't increase happiness". (Probably the happiness scores would top out before 10, though... Nobody thinks they're perfectly happy).

Will said...

Oh... I meant to say both of these problems would appear in cross-sectional comparisons and time-series.

Jason said...

Will, it's a sort of funny way of putting it but what you're saying is that when people are asked about happiness, they may automatically think about "relative happiness."

I agree that this is a good plausible explanation for the non-relationship. On the other hand, it doesn't explain why lottery winners don't stay happier than they were before hitting the jackpot. Perhaps they move to Beverly Hills where everyone is happier in absolute terms?

I do think consumption smoothing is *part* of the story but, regardless, I think people with more income *should* be happier.

Will said...

Lottery winners are explained by the psychology in Stumbling on Happiness. Lottery players expect being fantastically rich will make them fantastically happy. Their linear model is a good approximation in the small wealth neighborhood they've lived in (getting a raise made them about proportionally happy), but the true model has decreasing returns.

The psychologists say that happiness depends of the reference point and they say we are much more unhappy about losses compared to the reference point than gains. So the lottery winner's linear prediction is way away above the actual and they are highly disappointed.

In the lottery case, the winner has no chance to update her expectations. The lottery win is a one time big bang. In most cases, especially economic growth, agents have the opportunity to update their expectations and thus their reference points. Growth in incomes is usually slow enough to allow for learning.

So, I don't think the experience of lottery winners tells us much about overall happiness.

Jason said...

The fact that people are "not as happy as they expected" after winning the lottery doesn't explain why they do not remain happier than they were before winning the lottery.