Popular Posts

Caveat Emptor

The opinions expressed on this page are mine alone. Any similarities to the views of my employer are completely coincidental.

Tuesday, 2 September 2014

On Social Class: annus horribilis

Extended debates are boring, eventually even for those that  have a taste for them. They grow stale and take on the character of a pantomime routine:  'Oh yes you did’ ‘Oh no I didn’t’. Journals, quite rightly, tend to have a thrust, parry, counter rule with the original authors being allowed the last word.

In a world of blogs and post-publication commentary there is no such constraint, and only the exercise of self-control, or exhaustion can  limit the back and forth. But even I like to move on. I currently have a piece  with referees at Sociological Review which comments on the latest shenanigans of the GBCS boys and girls as they rapidly attempt to shift ground from the quicksands of class to what they think is the firm rock of 'elites'. I don't think they'll find it very comfortable reading, but I'm not going to reveal my hand here.

This is going to be my last extended blog post on Savage et al.’s 2013 GBCS paper and it’s occasioned by the OnlineFirst publication of their reply to their critics -On Social Class, Anno 2014 (OSCA2014). If you are already tired of this topic then don’t read any further. I understand and sympathize. 

So, having given fair warning, I hope what I have to say will be nuanced enough to satisfy both my readers and the rarefied tastes of Savage and his colleagues. I can’t promise to lessen the intensity of my hostility, but to be clear, it is not aimed at them personally. I don’t dislike any of them; I don’t even know half of them; what I dislike is what they  write and choose to disseminate. And I dislike it because I think, for reasons I will give, that large amounts of it are demonstrably nonsense. And what are academics for but to point out nonsense when they see it? We’re not supposed to be a mutual support club dedicated to propping up fragile egos or indeed fragile disciplines.

In the interests of brevity I've sacrificed style to content. I simply don't have the energy or interest to write another piece of joined up prose about the GBCS. Instead I've gone for a simpler format. I quote directly from OSCA2014  (all quotations in italics) and then follow up with my reply.

So here we go then.

1) What they say:
Whether these seven classes are analytically useful forms the focus of Bradley’s reflections, whilst Mills expresses doubts about the point of developing this classification in the first place (since the NS-SEC is a well-tested and validated measure of class already).

1) What I say:
I certainly have doubts about the point of developing the GBCS’s 7 class classification, for reasons that I’ve given elsewhere, but the mere existence of the NS-SEC is not one of them. This should really be obvious and it’s a big stretch beyond the facts to say otherwise. People are free to while away the hours inventing as many class schema as they like. The only sensible question to ask is: what are they good for? I’ve asked that question of the GBCS scheme and received no coherent answer. There’s really nothing more to say.

2) What they say:
…it is striking that none of the response articles makes any reference to the value of the ‘capitals, assets resources’ approach which informed our work. In particular, many of Mills’ questions about how we see the nature and scope of class analysis have been amply outlined in this literature, which he does not address (even though much of it is cited in our reference list).

2) What I say:
This is, to say the least, disingenuous in the extreme. Savage et al. know full well that in my original blog, in the longer critique I submitted to Sociology (which was desk rejected with the editorial comment: “…we do not want article length responses to Savage et al.”) and elsewhere I do take up the issue of the ‘capitals, assets, resources’ approach. Stripped of all the pretentious Bourdieusian verbiage there isn’t much to say, especially when, as now seems to be the case, the point has been conceded that ‘capitals’ is, in practice, just another word for resources. By the way Geoffrey Hodgson has just published an article in CJE which makes many points about the vacuous use of the word 'capital'  in sociology and economics much more clearly than I can.

3) What they say:
What is now at stake is the monopoly of the NS-SEC, institutionalised through the Office of National Statistics, to be the only public measure of class. We take this to be one of the reasons for the aggressive tone which Mills adopts in his response, which does not appear to be explicable as a reaction to the tone of our article which is respectful to different parties, including the proponents of NS-SEC, throughout.

3) What I say
Wow, that first sentence is portentous…and inaccurate. ONS has no such monopoly. Users of data – even ONS data - are free to code it up to any class schema they desire. Nobody is preventing them from doing so. As the national statistics provider ONS will obviously choose to use certain standard tools – like the NS-SEC  - that are the product of a carefully considered process of development and validation. A process, by the way, that involved extensive consultation with end users some of whom were even sociologists. This is what a national statistics provider does –develop measures of things where the details of how the measure was constructed are in the public domain. Sure, lots of people use the NS-SECs, why shouldn’t they? But they are not compelled to do so and the invocation of the word ‘monopoly’ is just a shabby rhetorical device employed in the service of what appears to be little more than self-aggrandisement. 

There is literally nothing at stake here and can’t be until such times as Savage et al. place in the public domain sufficient details about how to code data to their schema. That means, in case they are wondering, revealing the parameter estimates from their preferred latent profile model, something they so far seem remarkably reluctant to do. Potential users will also be interested in seeing the standard errors for these estimates (now I’ve got my tongue in my cheek for I know full well that there is no coherent way of producing them) so they can make a sensible evaluation of the reliability of the resulting classification. Oh, and of course, users will also have to have a data-set containing all of the variable, measured in exactly the same way, that go into Savage et al.’s statistical model.

Let’s face it chums, this isn’t going to happen. Quixote has tilted at his windmill and the NS-SECs are safe.

4)  What they say:
Firstly, and amply marked in Mills’ article, is the emphasis on (1) class as a ‘discrete’ variable, which needs to be delineated and differentiated from any other property with which it might be contingently affiliated. This endeavour to define class as a unitary variable can then lead to a broader project of empirically assessing its significance for other outcomes through using various kinds of multivariate model. It follows that, for this perspective, class needs to be differentiated from anything else with which it might be associated (status, gender, age, ethnicity, residential location, or whatever) so that it is stripped bare as a unitary phenomenon and its net significance registered. Much of Mills’ hostility to our article appears to be – no doubt deeply and genuinely held – bafflement that class could be anything other than a discrete, validated variable of this kind.
However, as we thought we made clear, our preferred definition of class is different to this. We are seeking a measure of class as (2) class formation. Here, the crystallisation of different properties renders a class as having a social existence over and above the different factors which make it up. It is in this sense that historians have been interested in classes, not as ‘pure’ variables stripped of contaminants, but as distinctive social formations.

4) What I say:
The important words here are ‘contingently affiliated’.  Let’s back up a little bit and make things concrete by talking about some examples. 

When a class schema was being created for the Oxford Social Mobility Survey – way back in the middle of the 1970s – an important strategic choice confronted the investigators. How should apprentices be treated? To understand this you have to appreciate that the class categories were constructed by combining (cross-classifying) occupational titles (actually OPCS occupational groups) and employment status. But what was the employment status of an apprentice? Were they simply an employee, like other employees? And, what for that matter was their occupation? Was an apprentice fitter the same occupation as a fitter? 

These are questions which have no obviously correct answer. They require a judgment that involves considering the purposes to which the schema would be put. In this case the study of social class mobility. Once the purpose is established it becomes a little easier to think about these issues. What would happen if we created a class schema in which an apprentice fitter automatically changed social class on completion of his apprenticeship? Well, given the number of people who at that time had served apprenticeships, one consequence would be that one would appear to observe large amounts of social class mobility. Clearly something does change when people completed their apprenticeship – for one thing they got a higher wage – but did they change social class? It is not immediately obvious that they did so. The OMS investigators decided that apprentices should be coded to the occupation/employment status combination of the job they were training for.

Another example. In the 1980s there was a lot of discussion about whether the proper unit for ‘class analysis’ was the household or the individual. Furthermore, if it was the household, should equal weight be given to both partners of a conjugal pair?  One proposal was to say yes in answer to this question and propose that a household’s class position changed when the work status or the class position of either a husband or a wife changed. Again there is no correct answer to the question, there are simply logical consequences of doing things one way rather than another. The joint household definition never really took off because of one of these consequences. In a world where women’s connection to the labour market is discontinuous households would change social classes as the hours of female labour market participation rose and fell generating an enormous amount of apparent social class mobility. Now, nobody would deny that levels of household welfare might rise and fall to mirror this, but is this the same as saying that people change their social classes? 

The point is this: conceptualisation implies choices and choices have consequences. Whether these consequences are acceptable or unacceptable depends upon the purpose of the conceptualisation. Not everything can be in focus at the same time. But if you don't spell out what the purpose of your conceptualisation is then you can't really be taken seriously because you are essentially just allowing yourself to say and see whatever you want  or whatever suits the needs of the current argument. We all cut ourselves some slack from time to time, but this approach is right off the lead!

Now what about 'class formation'?  Without further elaboration it is difficult to know exactly what Savage et al. have in mind. But let me make a guess that the spirit is captured in Edward Thompson's words:

" I do not see class as a 'structure', nor even as a 'category', but as something which in fact happens (and can be shown to have happened) in human relationships."

I also don't see it as a 'structure' , but I do see it as a category - or at least a concept - for the simple reason that this is what it is. It's a tool that  we use to organise our understanding of things that have happened to real people in real human relationships. Their experiences are real as are their relationships and they may or may not have characteristics which we choose to label class experiences or class relationships. Whether the individuals themselves do so or not is in one sense irrelevant - if workers die young of industrial diseases it matters not a jot to the sociologist of health whether the dying recognize themselves and their fellows as workers - and in another sense highly relevant - ie if they recognize themselves as a class or use a vocabulary of class this may have consequences in the real world for how they organize themselves to pursue their interests. This however has nothing whatsoever to do with whether of not some abstract category is 'stripped of contaminants' and the introduction of this argument by Savage et al is nothing but a red herring.

5) What they say
The alternative path was pursued by Goldthorpe who sought to pull the concept of class clearly apart from any reference to exploitation at all (e.g. Goldthorpe, 2000b; Goldthorpe and Marshall, 1992). Increasingly indebted to economists’ analysis of the nature of labour and employment contracts as a means of monitoring their workers, he specifically insists that there are no ‘zero-sum’ conflicts between classes.

5) What I say
Go back and read the relevant passages. For example here is Goldthorpe & Marshall (1992):

“Secondly, class analysis as we understand it implies no theory of class exploitation, according to which all class relations must be necessarily and exclusively antagonistic, and from which the objective basis for a critical economics and sociology can be directly obtained. Although exponents of class analysis in our sense would certainly see conflict as being inherent within class relations, this does not require them to adhere to a labour theory of value, or indeed any other doctrine entailing exploitation as understood in Marxist discourse. Nor must they suppose, as is suggested by Sørensen (1991:73) that what is to the advantage of one class must always and entirely be to the disadvantage of another. In fact much interest has of late centred on theoretical discussion of the conditions under which class relations may be better understood as a positive-sum (or negative-sum) rather than as a simple zero-sum game. And this interest has then been reflected in substantive studies in a concern with the part that may be played by class compromises in for example, labour relations or the development of national political economies and welfare states (cf. the papers collected in Goldthorpe (ed.) 1984).”

And here is Goldthorpe (2000b)

“I am puzzled by Sørensen’s use of the concept of exploitation—a word that I would myself gladly see disappear from the sociological lexicon—since it is not clear that it plays any vital role in his arguments. Its function in Marxist thought was to allow a fusion of normative and positive claims in a way that I would find unacceptable, as also, I believe, would Sørensen. For example, it is evident that he would regard the elimination of rents, and thus of exploitation, from labor markets as tending often to have highly negative human consequences. If invoking exploitation is no more than a way of flagging the presence of structurally opposed class interests that lead to zero-sum conflicts, then its use is innocuous but scarcely necessary.”

Professor Savage really must do his homework more carefully. Goldthorpe does not insist that “…there are no ‘zero sum’ conflicts between classes”. How he (Savage) could have come to persuade himself otherwise must remain a matter for speculation, but there is really no substitute for reading carefully what other people write, making a serious effort to understand it and most importantly representing it correctly. Clearly these are not fashionable practices or ones that will bring glittering prizes but I would have thought that it was required of any academic with a modicum of self-respect.

6) What they say
According to Mills, and to other critics such as Goldthorpe (2013), our findings are simply a ‘data dredging exercise’. Now, as we emphasised in our article, it is definitely the case that our analysis is as good as the construction of the variables, and hence ‘rubbish in, rubbish out’ definitely applies. Because our latent profile analysis has established seven classes out of the mix of measures which were used to construct them, it does not follow that we have defined seven ‘formed’ classes. For this to be the case, we need to reflect on whether they appear to make sociological sense, and whether they might be identifying a group which potentially has some coherence. Mills is of the view that the classes are partly an artefact of our variable construction. In our defence, let us firstly explain why we think our measures of capital are sociologically robust, and then go on to consider the plausibility of the kind of class groupings that the latent profile analysis produces.

6) What I say
I'm certainly critical of the way in which Savage et al. translate their concepts into measurements and it must be the case that what comes out of an inductive exercise is a function of what goes in - including the amount of information (N is arbitrary). My feeling is that any sociologist that couldn't construct a plausible sounding account of whatever comes out of a  sausage machine is probably in the wrong profession. We all know how easy it is to come up with just so stories. If you  had specified before hand what you expected to find and then demonstrated that you found it, I'd have taken you more seriously. But that's not what you did. Please can you tell me what the fine sounding phrase 'sociologically robust' means? Robust to what?

7) What they say
Similarly, our thinking about cultural capital is informed by extensive previous research, notably that reported in Culture, Class, Distinction (Bennett et al., 2009) and we are surprised that few of our critics, notably Mills, seem to have sought out this book to inspect more fully the underpinning of our thinking. We briefly repeat some especially salient points.

7) What I say
Vanity oh vanity! It is obviously inconceivable to Savage et al. that anyone could have read Bennett et al. (which I have, I even inflicted it on some of my students last year, though I’m not entirely sure they all enjoyed the experience) without finding much in it that illuminated anything terribly pertinent.  I certainly didn’t find their discussion of ‘capitals’ very persuasive or in fact coherent. How they can claim to know what I have or haven’t sought out is a bit of a mystery to me.

8) What they say
Mills’ argument is that the questions on cultural practices (for instance, a taste for certain kinds of music) conflate age and class, but he seems to assert an almost ‘naturalist’ view that being young or old necessarily imparts a pre-disposition to certain cultural practices.

8) What I say
Of course I say no such thing and this whole sentence is just a sly smear. I don’t know what a ‘naturalist’ account of cultural tastes would look like (do they mean to imply I think that cultural tastes are genetically programmed?) but the use of the term is obviously selected to produce a knee-jerk reaction of the boo-hiss type. 

It is a fact that tastes for certain types of music are strongly correlated with age. Bennett et al’s and Savage et al’s data clearly show this and it is no way surprising. So are tastes (or the capacity) for certain types of pastime – for example going to the gym, participating in sports, etc. So, to a lesser degree, are eating out choices and watching Disney movies(when you have young kids you don’t eat out as much as you used to and  when you do you tend to go to places that will make the kids happy, McDonald's, Pizza Express or the branch of Giraffe at the local Tesco).

 In my view the association with age is most straightforwardly described in terms of cohort, period or age differences. I simply can’t see what is to be gained from saying that someone who stops playing Sunday League football at age 40 because they’ve become  too slow, or that someone who takes advantage of the kids growing up to go out for a curry and a pint on Friday night, has changed social class. This is what my understanding of the logic of Savage et al’s approach implies. They can define the word ‘class’ in any way they want. But if they wish to be taken seriously they must then acknowledge the consequences of their definitional decisions wherever these may lead. If they lead to the original conception collapsing under the weight of absurdities then that should tell them that their conception wasn’t fit for purpose in the first place.

9) What they say
Secondly, Mills claims that the cultural tastes and practices revealed in our Figures 1 and 2 (Savage et al., 2013) are actually the product of the NS-SEC class divisions which we are supposed to ‘disdain’. We would never deny – and indeed have ourselves argued that – NS-SEC classes are associated with these patterns (as are income, educational qualifications, and other indicators of social hierarchy, see Bennett et al., 2009). However, it does not follow that occupational class is the best predictor of such cultural patterns, and indeed there is extensive research which we cited in our article (such as by Chan, Goldthorpe, as well as by ourselves) which argues otherwise. Our own comment, which he cites, was a discussion of these other sources.

9) What I say
Firstly I challenge Savage et al to show where exactly I say that cultural tastes and practices “…are actually the product of the NS-SEC class divisions”. They won’t be able to do that because the statement is a figment of their imagination. I hope they have the common courtesy to retract a claim that is blatantly false. Secondly they say( pp 222 of the original article):

 “..the schema[referring to the Goldthorpe class schema and derivatives thereof] has been shown to be of less use in explicating the wider cultural and social activities and identities (see generally Devine, 1998; Savage, 2000), which do not appear to be closely linked to people’s class position, as defined by the Goldthorpe class schema [my emphasis], and alternative schemas have been proposed to explain patterns of cultural consumption (Le Roux et al., 2008)”

Rather self-evidently they are not referring to discussion in the work of others as they only cite themselves. QED their new claim directly contradict their own words -  compare: “We would never deny …that – NS-SEC classes are associated with these patterns” versus  “…which do not appear to be closely linked to people’s class position”. It’s difficult to know how to have a rational conversation with people that insist on maintaining two contradictory positions and then appeal to one or the other as they find convenient.

10) What they say
But it is quite erroneous to see our classes as simply the product of age divisions, as in Mills’ claim that ‘(L)ife cycle plays a role in distinguishing what Savage et al. term the “elite” and the “established middle class”’. Because the elite are 11 years older than the established middle classes, he thinks this will explain the superior economic capital which the older ‘naturally’ accrue. But in fact, at least as far as household income is concerned, Goldthorpe and McKnight (2004) show that 57-year-olds in class 1 and 2 are actually likely to have marginally lower income than 46-year-olds.

10) What I say
Again Professor Savage needs to make a better job of his homework. Goldthorpe and McKnight (2004) say nothing at all about household income. What they examine is individual earnings from full-time employment. Household income and individual earnings from employment are not the same thing (obviously). So the citation is completely irrelevant. Honestly, I expect higher standards from my MSc students. 

11) What they say
It is a striking point that none of the four responses, with the partial exception of Bradley, reflects on whether or not the seven classes we delineated might be sociologically resonant.

11) What I say
I’ve no idea what ‘sociological resonance’ is: it doesn’t ring a bell. Perhaps it’s a close relation of that mainstay of rogues and charlatans – the appeal to the ‘sociological imagination’

12) What they say
 Mills implies at various points that it is easy to change cultural practices (for instance, in his comment about how changing Facebook friends might entail a change of class) and therefore that these are not sociologically salient.  Given the extent of sociological research which emphasises the powerful social structuring of cultural practices, it would be helpful for Mills to have produced sociological evidence for his alternative view. 

12) What I say
This is a remarkably silly comment. For starters I don't know what sociological saliency means in this context and secondly the validity of my argument doesn't depend on the citation of examples. It shouldn't take a genius to see that if you define a movement between social classes as being a movement between jobs that have different contractual employment conditions attached to them then a precondition for social class mobility is that somebody gives you a job. This is not something that is just a matter of will. No matter how much I might want to be an investment banker if I can't persuade a bank to hire me then wishing won't make it so. On the other hand if going to Abigail's Party means that I have to affect a liking for Demis Roussos I can easily go to Amazon & download a few MP3s so that I know what I'm letting myself in for. Becoming an investment banker involves surmounting a solid obstacle. Saying you like Demis Roussos or even learning to like him does not (though admittedly the latter is not without its challenges).

No comments: