Jonathan Falk writes:
This is an interesting story of heterogeneity of response, and an interesting story of “adversarial collaboration,” and an interesting PNAS piece. I need to read it again later this weekend, though, to see if the stats make sense.
The article in question, by Matthew Killingsworth, Daniel Kahneman, and Barbara Mellers, is called “Income and emotional well-being: A conflict resolved,” and it begins:
Do larger incomes make people happier? Two authors of the present paper have published contradictory answers. Using dichotomous questions about the preceding day, Kahneman and Deaton reported a flattening pattern: happiness increased steadily with log(income) up to a threshold and then plateaued. Using experience sampling with a continuous scale, Killingsworth reported a linear-log pattern in which average happiness rose consistently with log(income). We engaged in an adversarial collaboration to search for a coherent interpretation of both studies. A reanalysis of Killingsworth’s experienced sampling data confirmed the flattening pattern only for the least happy people. Happiness increases steadily with log(income) among happier people, and even accelerates in the happiest group. Complementary nonlinearities contribute to the overall linear-log relationship. . . .
I agree with Falk that the collaboration and evaluation of past published work is great, and I’m happy with the discussion, which is focused so strongly on data and measurement and how they map to conclusions. I don’t know why they call it “adversarial collaboration,” as I don’t see anything adversarial here. That’s a good thing! I’m glad they’re cooperating. Maybe they could just call it “collaboration from multiple perspectives” or something like that.
On the substance, I think the article has two main problems, both of which are exhibited by its very first line:
Do larger incomes make people happier?
Two problems here:
1. Determinism. The question, “Do larger incomes make people happier?”, does not admit variation. Larger incomes are gonna make some people happier in some settings.
2. Causal attribution. If I’m understanding correctly, the data being analyzed are cross-sectional; to put it colloquially, they’re looking at correlation, not causation.
3. Framing in terms of a null hypothesis. Neither of the two articles that motivated this work suggested a zero pattern.
Putting these together, the question, “Do larger incomes make people happier?”, would be more accurately written as, “How much happier are people with high incomes, compared to people with moderate incomes?”
Picky, Picky
You might say that I’m just being picky here; when they ask, “Do larger incomes make people happier?”, everybody knows they’re really talking about averages (not about “people” in general), that they’re talking about association (not about anything “making people happier”), and that they’re doing measurement, not answering a yes-or-no question.
And, sure, I’m a statistician. Being picky is my business. Guilty as charged.
But . . . I think my points 1, 2, 3 are relevant to the underlying questions of interest, and dismissing them as being picky would be a mistake.
Here’s why I say this.
First, the determinism and the null-hypothesis framing leads to a claim about, “Can money buy happiness?” We already know that money can buy some happiness, some of the time. The question, “Are richer people happier, on average?”, that’s not the same, and I think it’s a mistake to confuse one with the other.
Second, the sloppiness about causality ends up avoiding some important issues. Start with the question, “Do larger incomes make people happier?” There are many ways to have larger incomes, and these can have different effects.
One way to see this is to flip the question around and ask, “Do smaller incomes make people unhappier?” The funny thing is, based Kahneman’s earlier work on loss aversion, he’d probably say an emphatic Yes to that question. But we can also see that there are different ways to have a smaller income. You might choose to retire—or be forced to do so. You might get fired. Or you might take time off from work to take care of young children. Or maybe you’re just getting pulled by the tides of the national economy. All sorts of possibilities.
A common thread here is that it’s not necessarily the income causing the mood change; it’s that the change in income is happening along with other major events that can affect your mood. Indeed, it’s hard to imagine a big change in income that’s not associated with other big changes in your life.
Again, nothing wrong with looking at average associations of income and survey responses about happiness and life satisfaction. These average associations are interesting in their own right; no need to try to give them causal interpretations that they cannot bear.
Again, I like a lot of the above-linked paper. Within the context of the question, “How much happier are people with high incomes, compared to people with moderate incomes?”, they’re doing a clean, careful analysis, kinda like what my colleagues and I tried to do when reconciling different evaluations of the Millennium Villages Project, or as I tried to do when tracking down an iffy claim in political science. Starting with a discrepancy, getting into the details and figuring out what was going on, then stepping back and considering the larger implications: that’s what it’s all about.
“I don’t know why they call it “adversarial collaboration,” as I don’t see anything adversarial here. ”
The phrase adversarial collaboration does not have a compositional meaning. It refers to a collaborative investigation by researchers who have previously stated opposing opinions about the question to be studied. It is perhaps better parsed as a collaboration among people whose initial beliefs are in conflict.
Andrew has some good points (as always ;o) Maybe I am being Captain Obvious here. But, I am very troubled by attempts to measure things like happiness. I have been involved in studies of pain and depression. These mental barometers are extremely to difficult to measure.
Yeah, I’m not so sure about the utility of this either – as some kind of generic analysis.
I would guess that something like income inequality would affect the correlation between income and happiness. I’ll note that supposedly the happiest countries (while certainly skewed towards higher average income) are places that have good schools, good work/life balance, good social services, functional civil society, etc. Those things seem to me to be pretty tough to disaggregate.
It’s pretty hard to be happy when you are one trip to the ER/one illness/one “out-of-network” doctor away from bankruptcy, as the majority of Americans are.
And while the US economy is doing great (unemployment down, income and purchasing power up), the US housing market is a diaster (inadequate supply). Something like 50% of Americans would be living somewhere else if reasonably-priced housing were available. So the economny doesn’t feel great to people who can’t find a place to live they like.
So there are lots of confounding factors that make income nowhere near as important as an economist would like to think.
The more I think about it…
I was thinking along the lines of the interaction between wealth inequality and income when assessing the causality behind crim rates. It seems there’s good evidence that with increased income inequality, relative poverty plays a bigger causal role. I’d be surprised if there weren’t a similar interaction with happiness. Someone might be more “happy” happy at their income level of they I don’t have much exposure to people who have a shit-ton more moment than they. Thus, the incremental association between I come and happiness would change depending on context.
Relatedly, the whole concept seems pretty WEIRD (western, educated, industrialized, rich, democratic) focused to me. Sure, pretty much everyone over the world wants more money, but I’d guess the value placed on money (and hence it’s correlation with happiness) would vary across different cultures. That wouldn’t negate the averaging association across cultures, but it could significantly change the nature of what the association means.
Finally, I think that research shows that (on the West, at least) there’s a “U” shaped curve in happiness as people age (the young and the old tend to be more happy). That would be mostly reversed when you lool at income.
Many questions…
When will I remember to proof read (at least when I’m on my phone)?
Many typos but I think most can be reckoned with.
Joshua wrote:
“Someone might be more “happy” at their income level if they I don’t have much exposure to people who have a shit-ton more moment than they. ”
Yep. The “keeping up with the Joneses” problem. Old as the hills.
There were, at one point, a serious of incredibly stupid David Brooks (and/or friends???) editorials on how tough life was for folks only making US$400,000. Private schools, expensive cars/houses. It’s hard to make ends meet on US$400,000.
I’m rather amused at this thread for a completely off-topic reason. The Japanese strike me crazy over-the-top overconcerned with “shiawase” (happiness, of course). If it’s so damn imporant, just take drugs, I feel like screaming at them some times. And here we academics are being concerned about happiness.
Anyway, it seems like the wrong goal. And also to be seriously unquantifiable…
I’m not sure why Kahneman’s work on loss aversion implies that smaller incomes make people unhappier. His main point with regard to loss aversion is that people strongly dislike the prospect of negative changes from their reference point. If your income is small but steady, then according to him loss aversion shouldn’t kick in. Also, as far as I’m aware, the loss-aversion studies have all involved short-term decision-making. You can’t necessarily extrapolate from them to long-term happiness; it would be entirely consistent for Kahneman to say the reference point gradually adjusts over time.
Kevin:
The problem here is that the phrase, “Do smaller incomes make people unhappier?”, is ambiguous. I was taking the word “make” to imply a causal statement, i.e., “Does lowering a person’s income make that person unhappier?” You’re giving a comparative interpretation, i.e., “Are people with smaller incomes unhappier than people with larger incomes?”
I think this is a big problem, that in that article they’re conflating the causal question of within-person change with the question of comparisons between people, and writing as if they’re the same thing.
Lowering income could also come from a lot of causes. A person retires from being a corporate lawyer, buys a boat, and spends time deep sea fishing with their friends… Could massively improve their happiness! On the other hand, a successful construction business owner is forced to shut their business when they go bankrupt after an accident when their insurance carrier also goes under due to fraud…. gonna suck.
Every *way* in which people’s income is lowered matters. If it’s the mob coming around and extracting “protection” money from your small business… it’s a lot different than you proudly hand half of the business over to your children who take over most of the day to day operations… etc
Daniel: this paper studies only *employed US residents* between the *ages 18 and 65*. Income decline resulting from retirement isn’t relevant; nor is income decline – at least so far as I know – due to gang extortion likely to be a significant issue.
People retire early, have their income attached by an opponent in an ugly divorce, go bankrupt when their accountant runs off to Acapulco with their business funds, have sudden serious injuries preventing them from working…
I’m always amazed that people are so literal with my examples. The point of the examples is to get the imagination flowing. I think it works for some people but for others they focus in on the specifics of individual examples and miss the point.
Isn’t there a top to happiness? If someone handed me a million dollars on the day that the Cubs won the World Series, the million dollars would have had little effect on my happiness. It seems there is probably an asymptotic path to happiness. Going from starvation to adequate calories is a big step; going from hamburgers to steaks is a lesser step. Going from a ribeye to a New York strip is a small step.
Hmm. I don’t have much use for steak, but an occassional burger is fun. There’s some fun things Japanese, but the Japanese tend to put something in that messes it up. Cold buckwheat noodles (soba) and tempura is great, if and only if they leave off the damn nori. Chawanmushi is one of the greatest inventions in human history, if and only if they leave out the sea urchins. And the problem gets worse and worse the more upscale you go. Like they say here “Tade kuu mushi wa sukizuki” (“The bug munching on stinkweed goes “Tasty tasty””, and it’s also a weird novel by Tanizaki (“Some Prefer Nettles”).)
So soba (if I remember to order seiro), home cooked chawanmushi, and Indian bump up the happiness scale significantly. Generic Chinese in Tokyo is (IMHO) uninspired, but for a while, we had around the corner from us a home-cooked-style, hole-in-the-wall, the rest of the customers were construction workers Chinese restaurant that was off-scale great.
Here, not being hassled/not having hassles is big on the happiness scale. The electrical storm yesterday induced a power surge that blew one LED bulb and tripped the main circuit breaker. And burned out the water heater. (And, go figure, stopped a _battery operated_ clock.)
Hassle city. So I’m grumpier than usual. (It was actually a lovely storm. Loudest thunder I’ve ever heard (which was why the induced surge), dropped the temperature from 95 to 55 instantly, close to the heaviest rain we’ve seen. But the water heater is going to be serious money. Ouch.)
This was my immediate response, especially where they say there is “leveling” at the top and bottom. Especially in survey data with finite measures. Converting them to Z scores does not change the fact that there is an a fixed maximum.
“If someone handed me a million dollars on the day that the Cubs won the World Series, the million dollars would have had little effect on my happiness.”
How often has that happened to you so that you would *know* how happy that would make you? :)
Whatever you might really do under those circumstances, Indeed.com gives oncologists a base salary of $224K in the US. That’s 3x the median US income and solidly in the top 10% of the population. Supposing you had a 40-yr career at that salary relative to the median and earned 80% average market returns. You’d probably have a few million or more lying around, so its not surprising that a million wouldn’t impress you that much.
Interestingly though, while this study doesn’t include retired people, I doubt the happiness / income relationship would hold as people age. As one’s health declines in old age, one’s ability to do things also declines. Since no amount of money can save people from old age, money becomes less useful in creating happiness.
Chipmunk:
Good point regarding money and age. A lot of the value of money comes from reducing uncertainty and giving flexibility to your life, and once you’re getting on in years the money-related uncertainty can be less of an issue if you have a stable place to live and are not broke, and flexibility is less of an issue if you’re set in your ways and no longer have a family to support. At the low end, money still clearly has direct value when you’re old (for example, eating steak vs. cat food for dinner), but I agree that as we get older the value of money is less clear at the high end, when there’s less uncertainty about future life paths.
As I mentioned above, I believe there’s evidence suggesting a “u” shaped relationship between “happiness” and age; higher among the young, lower towards middle-age, higher again among the old.
That the inverse of the “money makes you happier” dynamic.
Just to say I think there’s prolly a ton o’ interaction effects, mediators/moderators, so as to make the identified correlation not very informative and the putative causal relationship pretty spurious.
Joshua:
Tthe possibly u-shaped relation between happiness and age came up in the blog a few years ago. The evidence isn’t so clear, and there’s a lot of controversy on the topic. I’m not saying there isn’t this u-shaped relationship, just that I’m not sure. Much seems to depend on how the question is asked and also on nonresponse, which is a particular concern if unhappy people are less likely to respond to the survey, and if that differential nonresponse varies systematically by age.
Andrew –
Interesting. Thanks. I’ll be sure to be even more unsure about this until further corrected.
Being a geezer, I often look up papers discussing some ailment I think I might have. Very frequently, the paper uses the language that “characteristic A” and “characteristic B” are linked. That sounds very vague — but I wonder how common it is, and in what ways it is interpreted by different people. Or is there some rigorous definition of “linked” in this context?
it probably means there is a vague positive associate between the two characteristics but not enough hint of a causal association to make a statement to that effect.
What does the (happiness, income) data “look” like?
1, The distribution of happiness depends on income.
2. The distribution is shifted to the right at larger incomes: but…
3. The magnitude of the shift depends on the place in the distribution; There are quantile effects.
4. The high quantiles (.9) of the happiness distribution shift more than the low quantiles (.1).
The original analysis missed these features of the data because it presumed that the data was well described by the relationship between expected happiness and income.
> it’s that the change in income is happening along with other major events that can affect your mood. Indeed, it’s hard to imagine a big change in income that’s not associated with other big changes in your life.
Along similar lines, seems to me it’s reasonable to ask about direction of causality. Can we rule out that there’s some mechanism whereby happier people make more money? I’m not saying it’s fully explanatory, but I’d think it could be a factor. Or maybe there are important moderator/mediator/interaction effects – say something like happier people are healthier, on average, and healthier people make more money?