In the comments on my marriage and happiness post last week Andrew Leigh and I differed on the link between mental distress and well-being. It started when Andrew pointed to this paper (pdf) to argue that, as he put it, ‘divorce makes you happier’ (compared to a bad marriage, that is).
On average, I am pretty sure that’s right. But the paper he cited did not use the standard tests for well-being, which ask people to rate themselves on a scale according to how happy they feel, or how satisfied they are with their lives. Instead, it used the General Health Questionnaire (GHQ), which
is used to detect psychiatric disorder in the general population and within community or non-psychiatric clinical settings such as primary care or general medical out-patients. It assesses the respondent’s current state and asks if that differs from his or her usual state. It is therefore sensitive to short-term psychiatric disorders but not to long-standing attributes of the respondent.
When I replied that I did not think the GHQ’s measures of distress could be easily extrapolated to a measure of happiness, Andrew’s response was that
I think of them as measuring the same underlying stuff. See for example Blanchflower & Oswald’s recent paper that ‘validates’ cross-country happiness measures by showing that they correlate negatively with hypertension.
Though it seems intuitively plausible that the GHQ and subjective well-being indicators measure the same ‘underlying stuff’, as with happiness and marriage this is an area of disagreement among happiness researchers. In Understanding Happiness: A Theory of Subjective Well-being, now quite old (1992) but still one of the most interesting books on the subject, Bruce Headey and Alex Wearing note that:
a large minority give themselves scores which are surprising either because they rate high on both well-being and psychological distress, or low on both.
In the National Health Survey of 2001 some people who were rated as having a high level of mental problems on the Kessler Psychological Distress Scale nevertheless rated themselves as ‘delighted’, ‘pleased’, or ‘mostly satisfied’ with their lives. There were more showing the expected pattern and rating their lives as ‘unhappy’ or ‘terrible’, but still we get the pattern observed by Headey and Wearing a decade earlier.
Given the nature of this group, we can’t rule out the possibility that some of them just aren’t thinking straight and give nonsense answers as a result. But there is a rational explanation as well, based on the difference between moods at various times and global assessments of how life is going as a whole. Just as it is possible to experience joyful moments when things are going badly generally, it is possible to experience mental distress when things are going ok. Headey and Wearing found that while life satisfaction and depression were strongly negatively correlated, life satisfaction and anxiety were only moderately negatively correlated. Anxiety is usually less serious than depression, and can come and go more easily, being triggered by particular situations but absent in other contexts.
In the case of people separating and divorcing, getting away from a bad relationship could greatly ease anxiety, without actually leading to happiness as we would normally think of it on the 0-10 subjective well-being scale (ie 7+). Indeed, the 2005 averages of the divorced and separated I reported last week were slightly below 7. Positive psychology, the self-help/clinical aspect of the study of subjective well-being, is built around this distinction. To achieve optimal happiness it is not enough to get rid of ill-being, the traditional focus of psychology, important as that is in itself. Real happiness is something more than just being free of anxiety, depression and other mental health problems.
Andrew Norton 
Out of the last six posts, four have been on a theme related to marriage. Is there a particular reason why this subject is of particular interest to you these days?
LikeLike
On a more serious note, I think there is also a need to differentiate between general happiness and specific happiness related to marriage. In my personal case the latter is quite low, but despite this I am in god spirits most of the time.
LikeLike
A littlestress does no harm but prolonged stress is not good
LikeLike
“Out of the last six posts, four have been on a theme related to marriage. Is there a particular reason why this subject is of particular interest to you these days?”
Two weddings to attend on one weekend put marriage in my mind and triggered the first post, and last night’s follow up to it was only incidentally about marriage (or to be more specific, divorce).
The gay marriage post was triggered by seeing a friend attack an acquaintance in The Australian, which I could write about as I had thought about the issue before. The liberal marriage post was a follow-up to something John Heard said in reply.
So nothing along the lines you are hinting at. It was just coincidence that two marriage-related topics appeared within a few days of each other, and both triggered comments that I wanted to reply to at greater length than I could in the comments box.
LikeLike
I have a good friend who I describe as my “gay, Christian, psychopathic ex-house mate with an IQ of 360.” Naturally, he seems mentally distressed most of the time. He is also exceedingly happy with his state!
___
BTW, the IQ estimate is not precise – just my way of saying that he’s several times smarter than me.
LikeLike
Andrew, we should probably engage the services of an econometrician who thinks more about latent variables than moi. But my intuitive response is to say that a negative correlation among a small group isn’t inconsistent with a positive correlation for the majority. For example, imagine two proxies for latent health: self-reported health (on a 1-10 scale), and body mass index. For most people, these will proxy the same thing. But for bodybuilders, they’ll be negatively correlated. Nonetheless, I think the general rule that self-reported health and BMI are both health proxies is a reasonable one.
LikeLike
Andrew – I don’t disagree that we will find many people on a spectrum from ill-being to well-being, and that people who are high on one will generally be low on the other.
But the overall point is I think an important one for this area of research, showing in the limited studies of ‘positive affect’ too – that there is an imperfect relationship between day-to-day (or hour-to-hour) moods and emotions and overall life assessment.
Also the psychological distress instruments simply aren’t designed to measure happiness. Of the 12 GHQ questions in the divorce paper only 2 deal with positive emotions, and even then at lukewarm level (happy enough all things considered). They are like a thermometer that can’t measure heat above 28 degrees.
LikeLike
Tom N, as someone once said “Better to be a Socrates discontented than a pig happy”.
LikeLike
AN, fair points. I like the thermometer analogy.
LikeLike
Yes, the mentally distressed can be happy. Of course it would depend upon the degree of distress the individual is going through.
Happiness may come in cycles throughout the day. Depression affects people more in the morning than afternoon or evening.
LikeLike