For the last few decades, research into social determinants of health and well-being seemed to tell a simple story: the poorer you are, or the lower your class, the worse your health and well-being is likely to be.
Figure 1 below shows a clear gradient of psychological well-being by class. The larger differences, compared with the professional class, are at the bottom of the hierarchy with those in unskilled jobs being the worst off, while the managerial class are more similar to professionals (the professionals are at zero and not shown on this chart as they serve as a comparison category in our analysis).
Figure 1: At face value, lower social classes have worse psychological well-being. Numbers are differences compared to professional class, after accounting for age, sex, and ethnicity. Health Survey for England
But what lies behind this gradient? What causes psychological well-being to be unequally distributed by social class? It is usually explained in terms of several “causal pathways”. One possible pathway is that people who in the lower classes not only have lower incomes but also probably have less stability in their incomes, making life more stressful. A second pathway is that people in lower classes may experience more job-related stress. Third, people in the lower classes tend to engage in more unhealthy behaviour (including smoking and lack of exercise – although this is not the case for drinking alcohol). Fourth is the “psychosocial” pathway – that being in a lower position in the social hierarchy reduces feelings of worth and self-respect, which in turn damage psychological well-being and health.
A new pathway
All these factors are probably important to differing degrees. However, in our new research on social class and psychological well-being, we suggest that employment status is another pathway that can explain the class gradient. We noticed that employment status is forgotten in many studies of well-being gradients.
Yet employment status is associated with far larger differences in well-being than class (see figure 2 – note that the scale is now much wider than in figure 1). We can see that the long-term ill average six points below those in employment, for the unemployed the difference is three points, with smaller differences for the early retired and those looking after the home or family.
Figure 2: People outside the labour market have worse psychological well-being. Numbers are differences compared to those in employment, after accounting for age, sex, ethnicity and income. Health Survey for England
We are making no claims of causality here. Of course the long-term ill have lower well-being than others. In fact, among people who are outside the labour market due to illness, it is becoming increasingly likely that this is due to mental rather than physical health reasons. For unemployment, the causal direction could run either way: people are more likely to lose their job if they have low psychological well-being, but unemployment also makes people feel miserable.
But what does employment status have to do with the class gradient? The answer is: quite a lot, as employment itself is unevenly distributed by social class. When we measure class for everyone, based on the last job if not currently in work, we can see (figure 3) that among professionals 91% are in the labour market and less than 1% are unable to work due to long term illness. Contrast this to the unskilled where just 55% are in employment and 8% have a long term illness.
Figure 3 – The chances of being in the labour market are far higher in the higher social classes Health Survey for England
So, given this association, it is perhaps not surprising that when we controlled for employment status in our analysis, the class gradient disappeared completely. In our analysis, we also delved deeper and asked whether pre-existing factors, such as childhood illness, might be able to explain these outcomes in adulthood.
We used a type of statistical model that compares the well-being of individuals at different points of time, when they have changed social class or their labour market status. Our results showed that changes in labour market status (such as moving from being employed to being ill, or returning to the labour market after having stayed at home due to caring responsibilities) are accompanied by changes in well-being. On the other hand, changes in social class make no difference at all.
Does this mean that class does not matter? Have sociologists been barking up the wrong tree all these years by talking about social class? We shouldn’t be too hasty. We have seen that social class is a key factor in determining access to the labour market, which has clear implications for psychological health. But the real culprit, as we’ve demonstrated, is employment status.
This opinion piece reflects the views of the author, and does not necessarily reflect the position of the Oxford Martin School or the University of Oxford. Any errors or omissions are those of the author.