Why is middle age so depressing?

I turn 40 this year. (Buy me something nice!) I’m sort of OK with it, mainly because the Covid-19 situation means that I don’t feel obligated to organise some enormous party or stag-do-like Lads’ Weekend Away. On the big day itself we’ll be in Center Parcs, because I’m a wild man. Still though: 40. It feels significant.

Here’s one reason why it feels significant. One of the most well-known findings in economics and psychology is the “U-shaped curve” of happiness. In short, it says that we’re happiest when we’re young, then there’s a decline, but later in life we cheer up again. The finding seems robust — the economist David Blanchflower has found a happiness curve in nearly 200 different countries, so it’s (probably) not just affluenza-induced mid-life crises in the rich West. And the point at which the average person’s wellbeing is recorded as lowest — male or female, developed world or developing — is in their 40s. 

As I said, it’s a well-known finding. And there’s a tendency — which I share — to ascribe a kind of destiny to it: as though we all sit on a great wellbeing rollercoaster, and go up and down in our cohorts, the 1980 babies all nearing the abyssal nadir at the same time, while their 1990 successors are just starting their downward run behind them.

I’m partly thinking about this because an interesting new study, by the Cambridge psychologist Dr Amy Orben and colleagues, finds something similar again: it suggests that “life satisfaction” (not quite the same thing as wellbeing) begins declining earlier, as young as 10, and starts coming back up around age 50. Another slightly unexpected thing it found was that although we tend to think of adolescent girls as suffering particularly badly, boys have a similar drop in wellbeing, just a couple of years later on average: Orben wonders if this is about the onset of puberty, and related brain development.

I am fascinated by happiness research. It seems to me that it might be the most important subject in the world; and yet it is also an absolute nightmare to study, because you’re trying to find answers that philosophers have argued about for three millennia with a five-item questionnaire.

Here’s what I mean by important. If you’re trying to cure cancer, why are you doing it if not, in the end, to make people happier? If you’re trying to build a better smartphone, why are you doing it if not, in the end, to make people happier? When a politician says they got into politics to “improve people’s lives”, what does that mean if not, in the end, to make people happier? 

Obviously I have just waded knee-deep into about six separate philosophical controversies there: what does it mean to be happy; is “the good life” about happiness, or about living a life of virtue, or a life that you are proud of, or what; is self-realisation and independence more important than happiness etc etc. 

But it seems to me that — on some, quite important, level — what matters is whether people are happy. If we cured all the world’s cancers, made the perfect smartphone and achieved the ideal form of representative government, but everyone was just as miserable after as they were before, then — did any of it matter? I think it is at least plausible that happiness, in some nebulous and hard-to-define but nonetheless real sense, is the most important thing in the universe.

And I don’t think it’s just me. The British government flirted with measuring wellbeing and using it as a national target, like GDP growth or inflation; the New Zealand government has actually gone further and done it. (At government level, everyone’s a utilitarian, I guess.)

Which means we have a problem. Or several problems, first of which is: what are we talking about? When psychologists use the term “happiness”, they seem to use it relatively interchangeably (in my experience) with “subjective wellbeing”, which is distinct from “life satisfaction”. 

Subjective wellbeing is how you’re feeling right now, in the moment; psychologists might measure this by asking someone to agree or disagree with the statement “I feel good about myself” or “I feel cheerful”. Life satisfaction is how you rate your life when you are asked to take stock of it; psychologists might measure this by asking people to agree or disagree with the statement “If I could live my life over, I would change almost nothing” or “I am satisfied with my life”. 

These measures are related, but distinct. To return to the U-shaped curve, most studies refer to wellbeing; the Orben et al paper is about life satisfaction. Dr Julia Rohrer, a psychologist at the University of Leipzig, gave me a couple of helpful examples of how they’re distinct.

First, often when people have children, their subjective wellbeing goes down: being a parent is hard, and you feel tired and harried and don’t get any time to yourself. But their life satisfaction goes up: they feel proud of their kids, and they add meaning to life. Conversely, she says, people who are unemployed often report reduced life satisfaction; but sometimes their subjective wellbeing actually goes up, because work is hard and not always all that fun.

(This isn’t true for everyone, obviously. We are always talking about averages across population.)

So we’ve defined it. But now the second problem is we have to measure it, and that’s difficult too. It’s not like measuring someone’s height, where there’s a clear and reliable system. You’re using a questionnaire to assess something internal and subjective.

There’s a quite deep philosophical assumption here, or two assumptions. First, that there is what Rohrer calls an “unobserved quantity”, something internal and real and true, that is our “happiness” or our “wellbeing”. And second, that this unobserved quantity somehow relates, in a reliable, predictable way, to what is going on when we say “I would rate my subjective wellbeing three out of 10.”

That doesn’t feel too implausible. But if you’re assessing someone’s happiness in Britain and someone’s happiness in China, are you getting at the same quantity? If you ask someone to rate from 0 to 10 how happy they are — does that work? Do the English word “happiness” and the Mandarin word “幸福” (xingfú, apparently), which Google tells me is the translation, really address exactly the same concept?

We don’t even need to go to China. Would the answer that a 16-year-old gives be directly comparable to the answer a 40-year-old gives? Or a 70-year-old? Are they reading those numbers off the same internal scoreboard? Whatever this “unobserved quantity” is, can we be sure that it is the same when you ask a child as when you ask her grandmother? “It can’t be the same construct across the life course,” is Rohrer’s answer. “Going from age 10 to age 80, it must be a completely different type of assessment.”

(Orben is entirely aware of this, I should say: “Measuring happiness and life satisfaction is extremely difficult and we interpret it differently throughout our lives,” she says. Her research is carried out in the full knowledge that it is difficult and uncertain and needs to be approached with caution. Psychology is the hardest science.)

So those are the problems with researching it. But there’s another layer of difficulty, which is that presumably we want to know about happiness in order to create more of it; but it seems that it is quite hard to make. Somewhere between 50% and 80% of the variation in happiness in the population is inherited with your genes; for many, if not most, of us, there is quite a hard limit to how much you can improve your happiness. 

That doesn’t mean that 50% of “your happiness” is in your genes (what would that mean?), rather that the differences between people’s reported happiness scores in a given society are at least 50% down to their different genetic make-up. The rest is traditionally divided between our “circumstances” — our material wealth, our upbringing, our social environment — and our decisions.

You can see what I mean when you look at the impact of life events — childbirth, divorce, marriage, unemployment, loss of a spouse — on reported life satisfaction. In many cases (not all; unemployment is the obvious exception), people’s scores change dramatically, but relatively quickly return to where they were. There really does seem to be a set point, and while you can be shifted from it, you have a sort of inertia which tends to bring you back to it. (That said, in the case of widowhood, both men and women seem to end up happier a few years after the death of their partner, which is frankly disturbing.)

It may not be that there is literally a happiness gene, says Rohrer: “The heritability estimates don’t tell you why happiness is heritable,” she says. “It could be because happiness is innate, but we know cognitive ability is highly heritable, which would affect your achievement, which would affect your income, which would affect your wellbeing. It could just be that you’re blessed with skills that society happens to value.”

On which point, it’s worth noting that although the one thing that everyone knows about happiness psychology is that money doesn’t make you happy, money does, in fact, make you happy. Dr Nick Brown, another psychologist, points out that there is in fact a strong correlation between GDP and happiness, and (on a personal level) between income and happiness. “There’s an old joke I saw on an office wall,” he says. “‘Money doesn’t make you happy. People with $10 million are no happier than people with $9 million.’ Once you get above the middle-class income in a median area of the US [probably around $75,000] it might not make much difference, but actually the intuitive feeling, that people with more money are happier, is right.” 

(Again, on average. For some people this will be more true; for others, less so.)

This all has quite profound implications, or so it seems to me, for government policy. If we take the position (as many do) that government should improve happiness, and it turns out that material wealth and genes account for most of the variation in happiness, then the policies that would imply might be very different from if individual decisions are especially important for happiness. This stuff is hard to research, but the findings that come out of it seem to plug directly into the very core of what we want policy to do.

The Orben paper, incidentally, found an impressively large decline in wellbeing in early adolescence. Orben herself wondered whether that was to do with how our brains develop: that happiness, or life satisfaction, is heavily linked to comparisons with others — if you’re saying you’re not satisfied with how your life is, presumably you’re thinking it could have gone otherwise; and the only comparisons you have are other people’s. And the ability to make those comparisons doesn’t turn up fully formed with your milk teeth.

In adolescence, she says, “you have intense social development, a theory of mind develops, you’re thinking about how people are thinking, your perspective develops; you can compare yourself to people around the world” — you go from being an unreflective child to a reasoning, reflecting young adult, and suddenly the question “are you satisfied with your life” has a meaning beyond where the next ice cream is coming from. She is interested in how much of a role social media plays in it all, although her previous work shows that most of the more doomy claims about social media and wellbeing are overstated.

What interests me, of course, is why the average person gets happier as they get older. There’s some suggestion that it’s a cognitive process — our brains start focusing on the good things that have happened, not the bad. Rohrer, though, says “I don’t think there’s a single explanation for the observed pattern. One possibility is that midlife is worst because of many obligations — careers tend to peak with the associated workload, kids still need to be taken care of, parents may suddenly require care as their health declines.” (Basically, what’s going on in this cartoon.) As you get older, those responsibilities tend to decline.

Also, she says, midlife is where you have to face up to the reality of what you’ve achieved: “aspirations and expectations are compared to the actual outcomes”, with time for significant change running out. As you get older again, she says, some research suggests “that people realise what’s important: they focus more on positive emotions, and on close relationships that are particularly satisfying, rather than meeting new people.” It’s also important to remember, she says, that the U-shape is an average, “only visible in aggregate”, and while large compared to a lot of psychological findings, is still only a small effect in terms of overall happiness.

To return to where we started: researching happiness is hard, but it does seem that the U-shaped curve describes some real phenomenon. So I approach my fifth decade with a certain trepidation. But it’s not destiny, not some vast fairground ride of joy and gloom: it’s not that everyone follows the same path: some people get happier, some people get sadder, people’s happiness jumps around. It’s all about averages. 

And, at least, I don’t have to have a party.