If you’re even slightly familiar with the last decade’s deluge of pop science books in psychology, you probably have heard of the phenomenon that sudden events and life events like marriage, death of a spouse, winning the lottery, etc. don’t matter that much for happiness. Instead, you have a set point that you bounce back to in a couple of years, if not even faster. This has been quoted at least in Gilbert’s Stumbling on Happiness, and in Kahneman’s Thinking Fast and Slow. Here’s a typical pattern:
Source: Kahneman & Krueger (2006). Developments in the Measurement of Subjective Well-Being. Journal of Economic Perspectives, 20(1), pp. 3-24.
They all claim the same thing: even if you become a quadriplegic – or win the lottery – this has no impact on your happiness in the long term. In a few months or years, you’ll adapt and be right back to your happiness set point.
However, it’s not like that.
A meta-analysis from 2012 Luhmann, Hofmann, Eid & Lucas in the Journal of Personality and Social Psychology goes through a swathe of research about the impact of life events. Crucially, they look at longitudinal studies, which in this case are much better than just cross-sectional designs. Anyway, technicalities aside, let’s dive right into the main findings.
Here’s a picture from the paper:
They all claim the same thing: even if you become a quadriplegic – or win the lottery – this has no impact on your happiness in the long term. In a few months or years, you’ll adapt and be right back to your happiness set point.
However, it’s not like that.
A meta-analysis from 2012 Luhmann, Hofmann, Eid & Lucas in the Journal of Personality and Social Psychology goes through a swathe of research about the impact of life events. Crucially, they look at longitudinal studies, which in this case are much better than just cross-sectional designs. Anyway, technicalities aside, let’s dive right into the main findings.
Here’s a picture from the paper:
Here, the event (childbirth) happens at time t=0. The points are effect sizes, which compare the difference in emotional or affective well-being (AWB) to the value at the event. The CWB; LS and CWB: RS reflect to effect sizes for life satisfaction and relationship satisfaction. In the middle, the black straight line is the estimated level of AWB before the event, while the dashed straight line is the same, but for life satisfaction. The curves are just log model estimates for how the effect sizes for AWB, RS and LS develop over time.
The crucial point is that, even after 100 months (9 years!) the life satisfaction still hasn’t returned to the baseline. Since the estimated level before the event is negative, we know that LS is typically lower before the event than at t=0. This makes sense, since having a child is an exciting experience, and creates a good sense of achievement for many. For the life satisfaction to reach that baseline of before anticipating a child, it would have to reach the dashed straight line. This would then mean it’s that much lower than at childbirth, ie. the same as before anticipation.
Similar graphs can be found for marriage, divorce, losing one’s job, and rehiring. For example, the marriage one looks pretty similar:
The crucial point is that, even after 100 months (9 years!) the life satisfaction still hasn’t returned to the baseline. Since the estimated level before the event is negative, we know that LS is typically lower before the event than at t=0. This makes sense, since having a child is an exciting experience, and creates a good sense of achievement for many. For the life satisfaction to reach that baseline of before anticipating a child, it would have to reach the dashed straight line. This would then mean it’s that much lower than at childbirth, ie. the same as before anticipation.
Similar graphs can be found for marriage, divorce, losing one’s job, and rehiring. For example, the marriage one looks pretty similar:
Once again, the story is similar. Even after 10 years, life satisfaction is still not at the baseline, but slightly above the EPL. Emotional well-being seems to be unaffected by marriage, though there are only five estimated effect sizes.
For me, this is shocking. The adaptation hypothesis seemed to fit in together with everything I had read about happiness. (Of course it did, since all books referenced the same phenomenon.) Now, by golly, it looks like losing your partner does in fact make you unhappier. Even if this might be “common sense”, it’s prudent to remember that the same meme has been all over the place. Even in books and talks that have appeared after the meta-analysis.
If you’re a research psychologist who specializes in happiness, this is probably no news. However, if you’re anybody else, chances are that the happiness adaptation meme has found its way to your mind and entrenched itself deep. It certainly did that for me. I mean, you keep seeing the same thing in every book, so it must be true! But like so many memes, this one if false too.
What are the implications? Well, for me, this definitely decreases my confidence on the whole in the happiness set point hypothesis. I used to think that the set point was probably generated through some interaction of genetics, early life experiences and social environment. I used to think that it was very robust to changes in your personal wealth, job situation etc. Now, I’m not so sure. It could still be that the set point hypothesis still holds. Maybe the set point is just not as robust as I used to think.
However, what the meta-analysis seems to imply is that the set point itself can be changed. If the life events can impact your set point over the course of several years, it makes more sense to talk about change in the set point, instead of lags of several years.
P.S. On a tangent, finding out about this article was interesting validation for reading outside my own field (commonly called procrastinating). Wellbeing psychology is generally interesting, but I’m definitely not an expert on it. If I tried to read the journals in the field, I’d never get any work done – and would suffocate myself with what are (to me) irrelevant papers. There’s just too much stuff. But how do you separate the wheat from the chaff? Blogs can help: this gold nugget came through a psych/science blog, which had mentioned the finding (thanks, Scott). Call me out on the N=1 if you want, but now I feel again that this blog-reading is useful (and not just pointless PhD procrastination).
For me, this is shocking. The adaptation hypothesis seemed to fit in together with everything I had read about happiness. (Of course it did, since all books referenced the same phenomenon.) Now, by golly, it looks like losing your partner does in fact make you unhappier. Even if this might be “common sense”, it’s prudent to remember that the same meme has been all over the place. Even in books and talks that have appeared after the meta-analysis.
If you’re a research psychologist who specializes in happiness, this is probably no news. However, if you’re anybody else, chances are that the happiness adaptation meme has found its way to your mind and entrenched itself deep. It certainly did that for me. I mean, you keep seeing the same thing in every book, so it must be true! But like so many memes, this one if false too.
What are the implications? Well, for me, this definitely decreases my confidence on the whole in the happiness set point hypothesis. I used to think that the set point was probably generated through some interaction of genetics, early life experiences and social environment. I used to think that it was very robust to changes in your personal wealth, job situation etc. Now, I’m not so sure. It could still be that the set point hypothesis still holds. Maybe the set point is just not as robust as I used to think.
However, what the meta-analysis seems to imply is that the set point itself can be changed. If the life events can impact your set point over the course of several years, it makes more sense to talk about change in the set point, instead of lags of several years.
P.S. On a tangent, finding out about this article was interesting validation for reading outside my own field (commonly called procrastinating). Wellbeing psychology is generally interesting, but I’m definitely not an expert on it. If I tried to read the journals in the field, I’d never get any work done – and would suffocate myself with what are (to me) irrelevant papers. There’s just too much stuff. But how do you separate the wheat from the chaff? Blogs can help: this gold nugget came through a psych/science blog, which had mentioned the finding (thanks, Scott). Call me out on the N=1 if you want, but now I feel again that this blog-reading is useful (and not just pointless PhD procrastination).
RSS Feed
