Harry over at Crooked Timber draws my attention for the first time to a much talked about book by Michael Marmot, The Status Syndrome. By way in introduction, he lets former CT guest blogger Bill Gardner introduce us to the books arguments. Here’s Harry’s shorter Gardner:
Marmot�s fundamental empirical finding is that there is a social gradient in health, that is, when you group people according to their places in social hierarchies, you find better health and greater longevity in each successively higher class. His classic work studied the British civil service. He placed civil servants into four grades: administrators who set policy, executives who carry it out, clerical staff, and support personnel. There was a four-fold greater mortality rate from ages 40-64 for support personnel versus administrators. This was a large effect, and much larger than the difference in mortality rates related to conventional measures of social class.
…
Marmot argues that the principal explanation for the status syndrome is not relative income, not higher rates of health-risk behaviors among the lower classes, and not status-related differences in genes. Income, heath-related behavior, and genes are all important determinants of health, but their effects are largely independent of the effect of your place in the social hierarchy and only partially explain the social gradient. What matters is autonomy:for people above a certain material threshold of well-being, another sort of well-being is central. Autonomy � how much control you have over your life � and the opportunities you have for full social engagement and participation are crucial for health, well-being, and longevity. It is inequality in these that plays a big part in the social gradient in health. Degrees in control and participation underlie the status syndrome.
It is true that the finding of a more accute differential in mortality rates is interesting and in need of explanation. Nevertheless, I’m not quite sure what Marmot’s thesis about autonomy is supposed to be. (I’ve not read the book, so this post is about ideas, not the specifics of that work.)
First, I find curious the claim that “Income, heath-related behavior, and genes are all important determinants of health, but their effects are largely independent of the effect of your place in the social hierarchy and only partially explain the social gradient.” In order to understand what this means, we need to know what Gardner (and, presumably, Marmot) means by one’s “place in the social hierarchy.” Yet unless Marmot is working with technically stipulative definitions, surely one’s place in the social hierarchy is (at least partially) constituted by one’s income and opportunities for, and access to, health care and education. So how could it be true that the effects of social hierarchy are “largely independent” of the effects of certain indicies of one’s place is that very hierarchy?
Moreover, I would have thought that any account of autonomy worth its salt would have seen a clear connection between “how much control you have over your life” and your income, health, and education. Why think there is any competition between an explanation that makes one’s health dependent upon one’s opportunities fully to engage and paticipate in valued social institutions and practices, and an explanation that makes health dependent upon income and “health-risk behavior”? That’s just a silly, false dichotomy.
There is, however, a way of interpreting Gardner’s account of Marmot’s view that would be interesting. If somehow the data showed that what mattered most was the amount of autonomy one has on the job, then we would have an interesting proposal. Such a view would say that it is not social hierarchies at large that matter, but rather more local, professional hierarchies. Perhaps, as is likely, much more stress is placed upon support staff than is placed upon executives, and perhaps this is the sort of stress that cannot be mitigated by more money or improved social interaction with fellow citizens outside of one’s workplace. I think that something like this is probably true, and that it points to larger problems that arise in “developed”, efficient, and (therefore) highly stratified societies. But such a view would not make “your place in the social hierarchy” the salient causal factor, and thus cannot be Marmot’s view (as Gardner describes it).
Bill Gardner | 16-Dec-04 at 10:11 am | Permalink
Paul, I very much appreciate your comment. You raise a good question:
‘First, I find curious the claim that “Income, heath-related behavior, and genes are all important determinants of health, but their effects are largely independent of the effect of your place in the social hierarchy and only partially explain the social gradient.” In order to understand what this means, we need to know what Gardner (and, presumably, Marmot) means by one’s “place in the social hierarchy.” Yet unless Marmot is working with technically stipulative definitions, surely one’s place in the social hierarchy is (at least partially) constituted by one’s income and opportunities for, and access to, health care and education. So how could it be true that the effects of social hierarchy are “largely independent” of the effects of certain indicies of one’s place is that very hierarchy?’
By ‘place in the hierarchy,’ I meant the organizational grade described in the previous paragraph (‘administrators who set policy, executives who carry it out, clerical staff, and support personnel’). Of course, this grade will be correlated with income, but not perfectly. What Marmot believes that he has shown is that one’s place in an organizational hierarchy — which he thinks is a proxy for autonomy — is more strongly associated with health than income. Concerning access to health care, Marmot would point to social gradient related differences in health even within persons in the upper classes. He believes that these persons have similar access to health care. Marmot may be wrong in these assertions, but if so it is an empirical, not a logical problem.
In summary, you are right that there is ‘a clear connection between “how much control you have over your life” and your income, health, and education.’ However, these things are not identical, even though they may be causally related. For example, having more income usually helps you gain control over your life, but not always. Years ago, I moved from research academia to an administrative position in a university medical center. The job was prestigious and very well paid. I lasted a year, largely because I hated the loss of control over my time and activities in a corporate environment.
Paul | 16-Dec-04 at 12:29 pm | Permalink
Bill, thanks for stopping by and adding to the goodness of our site! Your response to my post is very helpful. I do still have some questions though. (I imagine that at some point I’ll become bothersome and you’ll have to tell me just to read the darn book. Don’t hesitate to do that sooner rather than later..)
By ‘place in the hierarchy,’ I meant the organizational grade described in the previous paragraph (‘administrators who set policy, executives who carry it out, clerical staff, and support personnel’). Of course, this grade will be correlated with income, but not perfectly.
Here you have omitted “social” for the original quotation, and it is that term that raised the flags for me. For later in the post you draw the tentative conclusion that “the opportunities you have for full social engagement and participation are crucial for health, well-being, and longevity.” Since you seem to place the focus on social aspects of life, I thought you were trying to make the (surely plausible) point that the quality of one’s interactions with others–e.g. fellow citizens in political debate or legal aliens working the counter at the local supermarket–matters a whole lot to well-being. (Whether it bears on something as physiological as mortality is a more difficult, and very interesting, question.) My worry was that Marmot’s evidence and conclusions about work-place was not satisfactorily connected to the aspects of social interaction at large that you seemed to be drawing attention to.
One of my main worries about the ideas in your post was that issues of social status were being confused with issues of work-place control and work-related stress. The worry arises for me because my current academic work in part concerns the duties that may arise with the existence of hierarchies of social status which may have little or no effects on individual well-being, or whose only effects are to create social situations where some citizens feel obliged to show humiliating deference to others. Because “least” is a comparative term, the door is open for the worst off to be very well off indeed–e.g., they may be very healthy. Nonetheless, there may be reasons to change a social arrangement like this, tho they would have little to do with one’s physical or mental health. (To you credit you recognize all of this, when you say that “the findings shows that inequality is not just unfair, but harmful.”)
Because your post placed so much emphasis on health–”lack of control over our workplace doesn’t just offend our sense of what’s right, it kills us.”–I wanted to make sure that “social status” was not being defined too narrowly. I also wanted to make sure that Marmot’s theses about status were not taken to exhaust the moral questions about status that arise. If certain aspects of social hierarchies cause serious health problems, that is something we should be grateful to know. And I am indeed grateful for your posts on the book.
Bill Gardner | 16-Dec-04 at 3:48 pm | Permalink
Paul: I agree that the distinctions that you raise matter, and matter a lot. I do not think — nor does Marmot think — that it is only the workplace that matters. I agree that “the quality of one’s interactions with others–e.g. fellow citizens in political debate or legal aliens working the counter at the local supermarket–matters a whole lot to well-being.” The quality of interactions with one’s spouse and children, etc., are also central. However, the workplace is particularly important, in part because so much coercion occurs there, and because threats to employment have such a direct effect on our well-being. I address your remarks further at http://childhealth.typepad.com/maternal_child_health/2004/12/marmot_the_evid.html. Thanks again.
Paul | 16-Dec-04 at 5:28 pm | Permalink
I’ve added further comments over at Bill’s site.