Forbes post, “Why Life Satisfaction Isn’t Necessarily ‘U-Shaped’ After All”

Originally published at Forbes.com on December 6, 2020.

 

Happiness, experts say, is U-shaped: generally speaking, we are happy/full of life satisfaction as young adults but, as we reach middle age, we become less satisfied, with a trough in one’s early 50s; from this trough we rebound to ever-increasing satisfaction levels as we age. It’s remarkable, really, considering the physical infirmities we face, plus financial worries, loss of loved ones, and more. What explains this? We become wiser and we are able to see all of life’s ups and downs with a greater sense of perspective.

But what if that’s not true?

new working paper by Peter Hudomiet, Michael D. Hurd and Susann Rohwedder, researchers at RAND Corporation, suggests an entirely different answer: older individuals have greater life satisfaction because the less-satisfied folk have been weeded-out. And by “weeded-out” I mean that they’re dead or otherwise unable to reply, because the likelihood of dying is greater for those who have less life satisfaction. When they apply calculations to try to strip out this impact, the effect is dramatic: rather than life satisfaction climbing steadily from the mid-50s to early 70s, then remaining steady, they see a steady drop from the early 70s as people age.

Here are the three key graphs (used with permission):

First, life satisfaction plotted by age without any special adjustments:

Second, the difference in mortality between the satisfied and the unsatisfied:

And, third, the same life satisfaction graph, adjusted to take into account the impact of the disproportionality of deaths:

In this graph, the blue line represents the unadjusted outputs from their calculations, the orange line is smoothed, and the grey line adds in demographic, labor market and health controls, to strip out the impact of, for example, people in poor health being less satisfied and try to isolate the impact solely of age.

Here are the details on this calculation.

The data they use for their analysis comes from the Health and Retirement Study (HRS), a long-running survey of individuals age 51 and older at the University of Michigan, sponsored by the National Institute on Aging. It is a longitudinal study; that is, it surveys the same group of people every two years in order to see how their responses change over time, adding in new “refresher cohorts” to keep the survey going. The survey asks about many topics, including income, health, housing, and the like, and in 2008, the survey also began to ask life satisfaction, on a scale of 1 to 5 (”not at all satisfied” to “completely satisfied”).

One simple way of analyzing the data is to look at how life satisfaction ratings vary based on survey participants’ characteristics. The average reported life satisfaction of those between ages 65 – 74 is 3.91, just slightly below “4 – very satisfied.” But those who rate their health as “poor” average out to 3.13, or not much more than “3 – somewhat satisfied,” and those who rate their health as “excellent” average to 4.34. Those who have 2 or more ADL (activities of daily living) limitations some out to an average of 3.32 vs. 3.97 for those with no such limits. Those who are in the poorest quarter of the survey group come out to 3.7 vs. 4.07 for the wealthiest quarter. (See the bottom of this article for the full table; this table and the following graphs are used with permission.)

But here’s the statistic that throws a monkey-wrench into the data:

“On average, the 2-year mortality rate [that is, from one survey round to the next] is 4.4% among those who are very or completely satisfied with their lives, while it is 7.3% (or 66% higher) among those who are not or somewhat satisfied with their lives.”

As a result, “those who are more satisfied with their lives live longer and make up a larger fraction of the sample at older ages.”

Now, this does not say that being pessimistic about one’s life causes one to be more likely to die. Nor does it say that this pessimism is justified by being in ill-health and at risk of dying. But this statistical connection, as well as further analysis of survey drop-outs for other reasons (such as dementia) is the basis for a regression analysis which results in the graph above.

What’s more, the original “inventor” of the concept of the life satisfaction curve, David Blanchflower, published a follow-up study just after this one. One of their key concepts is the notion of using “controls” to try to identify changes in life satisfaction solely due to age rather than changes in income over one’s lifetime, for example, or other factors, and there has been extensive debate about whether or to what degree this is appropriate, given that the reality of any individual’s life experience is that one does experience changes in marital and family status, employment status, and the like. Having received pushback for this concept, they defend it but also insist that the U-shape holds regardless of whether “controls” are used or not. At the same time, Blanchflower is quite insistent that the “U” is universal across cultures, though (see my prior article on the topic) it really seems to require quite some effort to make this U appear outside the Anglosphere, which is all the more interesting in light of the John Henrich “WEIRDest people” contention (see my October article) that various traits that had been viewed by psychologists as universally-generalizable are really quite distinctive to Western cultures and, more distinctively, the United States.

But here’s the fundamental question: why does it matter?

On an individual level, to believe that there is a trough and a rebound offers hope for those stuck in a midlife rut. It’s a form of self-help, the adult version of the “it gets better” campaign for teenagers.

On a societal level, the recognition of a drop in life satisfaction for the middle-aged might be explained, by someone with the perspective of the upper-middle class, as the result of dissatisfaction with a stagnating career, failure to achieve the corner office, the challenge of shepherding kids into college, and the like. In fact, when I wrote about the topic two years ago, that’s how the material I read generally presented the issue. But Blanchflower’s new paper recognizes greater stakes: “These dips in well-being are associated with higher levels of depression, including chronic depression, difficulty sleeping, and even suicide. In the U.S., deaths of despair are most likely to occur in the middle-aged years, and the patterns are robustly associated with unhappiness and stress. Across countries chronic depression and suicide rates peak in midlife.” (In the United States, among men, this is not true; men over 75 have the highest suicide rate.)

And what of the decline in life satisfaction among the elderly?

The premise that the elderly become increasingly satisfied with their lives as they age is a very appealing one, not just because it provides hope for us individually as we age. It serves as confirmation of a more fundamental belief, that the elderly are a source of wisdom and perspective on life. Although it is Asian cultures which are particularly known for veneration of the elderly, the importance of caring for those in need is just as much a moral imperative in Western societies, even if without the same sense of “veneration” or of valuing them to a greater degree than others in need. Consider, after all, that the evening news likes to feature stories of oldsters running marathons or competing in triathlons or even just having a sunny outlook on life; no one likes to think of the grumpy grandmother or grandmother from one’s childhood as representative of “old age.” In this respect, “old folks are more satisfied with life” provided an easy to make the elderly more “venerable.” Hudomiet’s research might force us to think a bit harder.

 

Full table of impact of demographic characteristics on life satisfaction:

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Forbes post, “Where Are All The Happy Retirees?”

Originally published at Forbes.com on November 1, 2018.

 

Readers, have you heard this nugget of wisdom before?  There’s something innate in us as humans that hardwires us to be comparatively satisfied with life as young adults, lose some of that satisfaction as we wend our way through adulthood, bottoming out during the “mid-life crisis,” but then experience renewed satisfaction as we reach retirement years.

Here’s the Washington Postin 2017, citing a new scholarly analysis:

Happiness, those surveys show, follow a generalized U-shape over the course of a life: People report high degrees of happiness in their late teens and early 20s. But as the years roll by, people become more and more miserable, hitting a nadir in life satisfaction sometime around the early 50s. Happiness rebounds from there into old age and retirement. . . .

These similarities [among various studies] are even more remarkable given the differences in the underlying surveys, which were administered in different countries. They include the General Social Survey (54,000 American respondents), the European Social Survey (316,000 respondents in 32 European countries), the Understanding Society survey (416,000 respondents in Great Britain) and others. . . .

“There is much evidence,” the paper’s authors conclude, “that humans experience a midlife psychological ‘low.'”

There’s even a new book out, The Happiness Curve, by Jonathan Rauch, which cites extensive studies and shares individual stories of people reaching midlife and feeling a vague sense of dissatisfaction, offering readers in that midlife slump hope that they aren’t alone, that it’s a natural stage of life, and that their perception will just as naturally improve over time.

What’s more, this curve extends to a multitude of countries, though the curve itself is curve-ier in some countries, and comparatively flatter in others, according to Rauch’s book and according to an analysis from 2016 which looks at a total of 46 countries, in the form of (smoothed) curves of happiness levels and the age at which happiness bottoms out before growing again.  In some countries, such as Australia, Cyprus, Estonia, Greece, Ireland, Serbia, Slovenia, and Spain, and the U.S., curve is very pronounced, with a bottoming-out age generally in the 40s or early 50s.  In others, the “curve” is so flat or simply just downward-sloping, and the bottoming-out age so late, that it seems a bit of a stretch to call it a “U”; these include Austria (age 63.29), Finland (58.09), India (54.27), and Russia (81.57!).

But what’s noteworthy is that this study, in order to create these charts, does not simply take the raw data but rather adjusts it, controlling for “age, marital status, gender, employment, education and household income in international dollars.”  The aim, as Rauch discusses in his book as well, is to somehow derive a “pure” impact of aging alone.

But what happens when you don’t apply this analysis based on controlling for these factors but look at real people in their real lives?  The results look quite different.

Here’s a comparison of men in four regions:  the Anglosphere (the US, Canada, Australia, and New Zealand), Western/Southern Europe, the former Warsaw Pact countries (Central/Eastern Europe and the former USSR), and Latin America.

Life Satisfaction of Men in 4 regions, from 2015 World Happiness report data... [+] (http://worldhappiness.report/ed/2015/)

own chart

And here are the women in those countries:

Life satisfaction of women in 4 regions

own graph

in both cases based on data downloaded from the 2015 edition of the World Happiness Report.

Here’s the same data updated to 2018 — that is, using data from 2015 – 2017, from the working paper “Happiness at Different Ages: The Social Context Matters,” by John F. Helliwell, Max B. Norton, Haifang Huang, and Shun Wang (used with permission).

2018 Cantril ladder graphs, 4 regions, men and women.

Helliwell et al., used with permission

The data in all these cases is based on the “Cantril ladder,” a question which very simply asks poll participants to imagine “the best possible life for you” and the “worst possible life for you” and rank how they view their current life situation on a scale of 0 to 10.

(The report also provides data for Africa, the Middle East, and Asia, for a total of 9 regions, though I’m focusing on Europe, the Anglosphere, and the Americas as regions with some possible cultural similarity.  Among the Asian regions, East Asia has a similar, and actually more pronounced uptick, but I presume their strong value of revering elders is a factor.)

It is immediately clear that even in this three year span, patterns have changed in such places as NA/ANZ (perhaps due to reactions to Trump, Black Lives Matter, etc., in the U.S.?) and Western Europe (perhaps due to the impact of the mass influx of migrants starting in summer 2015?).  But what’s noteworthy is that, of these regions, only the Anglosphere region shows a pronounced U, with Western Europe and Latin America only showing a very small upward rise at the very oldest ages, hardly enough to qualify as U-shaped according to the definitions.

What’s going on here?

What happened to the U shape?  This is the puzzle.

For the former East Bloc, various sources provide the explanation that the consistently downward slope is simply the result of the misery of Putinism, and the need to adjust to the dramatic changes, albeit several decades ago by now, of a postcommunist world, having a harsher impact on older vs. younger individuals (though note that the region includes such countries as Poland and Hungary, now part of the E.U.).  It’s certainly the case that in Russia, alcoholism heavily afflicts the older generation. Perhaps the economic crises that various Latin American countries have experienced impact those residents similarly.

But no such explanation seems to fit for Western Europe, especially based on pre-2015 data, where, if anything, reports are that older folk, with secure pensions, are better off than young people stymied by a high unemployment rate or working on a contract basis without the cushy job guarantees of older workers.  It’s younger people who are living at home, unable to start families, jealous of the older generations.

And if life satisfaction in Western Europe differed from the Anglsophere in that it was steady or climbing, there’d be an easy story to tell, and it’d go like this:  “Americans have an excessive degree of ambition and desire for achievement, so that the first part of their life is the story of the attempt to attain their life goals, midlife is when they realize that they will not attain these goals, and in their later adulthood years, they have acquired the wisdom to understand that that’s not how life works.  But Europeans don’t have the same drive towards achievement and recognition, so they don’t have the same crash when this doesn’t happen.  After all, the Germans don’t even have a word for ‘midlife crisis’ except the borrowed English word.”

But Western Europeans have the same drop; they simply don’t recover.

Why?  I don’t know, and the literature I’ve read doesn’t know.  But that’s not going to stop me from sharing a few theories.

Theory One:  what if the poor economy, despite secure pensions, is impacting older people because they are watching their children struggle?  This would suggest that life satisfaction at older ages is connected up with seeing the younger generation prosper, rather than just with one’s own personal well-being.

Theory Two:  what if it’s Europe’s low birth rates that make a difference?  After all, the birth rate has been low for years and years, peaking in the Eurozone at 2.733 in 1964, at which point it began dropping steadily, to 2.393 in 1970, 2.023 in 1975, 1.774 in 1980, 1.534 in 1990, and bottoming out at 1.383 in 1995.  And, yes, if you were in your prime childbearing years, say, age 30, in 1995, you’d be at that age now when, in the Anglosphere, on average, you’d be experiencing a rebound in your life satisfaction.  But if having children is both, in part, a driver of midlife stress, and a source of postmidlife satisfaction, then the low birth rate (in addition to such factors as are causing it in the first place) could be a clue, too.

Theory Three:  what if it’s differences in religiosity?  Pew polling reports that on multiple measures, Europeans report lower levels of religious practice.  11% of Western Europeans polled in 2017 said that “religion is very important in their lives,” compared to 53% of Americans.  22% vs. 50% “attend religious services at least monthly” and 11% vs. 55% say they pray daily.  Could a religious orientation make the difference in that uptick, all other things being equal?  (Note that sub-Saharan Africa has a low and declining life satisfaction, despite its greater religious practice, but that’s hardly apples-to-apples.)

And Theory Four:  what about different “locus of control” perceptions?  This refers to differences in attitudes about the degree to which your life is what you make of it, vs. your fate largely being out of your control.  For reference, I pulled out the results from an older book (1998) on cultural differences, Riding the Waves of Culture, by Fons Trompenaars and Charles Hampden-Turner.  Survey data from 50-some countries asking people whether they “believe what happens to them is their own doing” showed a range from 33% in Venezuela to 88% in Israel and Uruguay.  The third-ranked country was Norway, even though the other Nordics were in the middle of the pack.  And countries #4 through #7 were exactly the grouping that has the “real-world” U-shape:  the USA, Australia, New Zealand, and Canada.  What’s more, the UK and Ireland (part of the Europe grouping) tie for #8 along with Switzerland.

If the uptick after midlife is at least in part a matter of being able to say, “yes, I made something of my life,” then it stands to reason that having a strong personal sense that one’s own decisions and actions have a real impact on one’s life is a necessary ingredient in feeling a sense of satisfaction afterwards.

This is all speculation, of course, but you’ll notice that none of these explanations have anything to do with the quality of state or private retirement systems, because “successful” aging is about much more than this.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.