The Middle Classes: Workers, not Owners

by Hugo Radice

Viewing classes simply as social groups based on occupation, income, status and lifestyle avoids the need to identify and explain class relations under capitalism.

First published: 20 February, 2014 | Category: Class, Economy

In a recent Guardian article, Paul Mason discusses the rapid growth of a global ‘new middle class’ of educated, skilled and mostly non-manual workers as a result of globalisation. He argues that while rising incomes have lifted them out of desperate poverty, they now fight on issues such as corruption, democracy, urban infrastructure and healthcare.

The article was informative and refreshing in many ways, especially in situating the question of class globally rather than nationally, and in taking account of the universal shift from manual to non-manual work. But it also revealed the continuing confusion that bedevils public discussion about class – both here and in the rising economies of the global south. In particular, what do we mean when we speak of the middle class, ‘old’ or ‘new’, or indeed the plural ‘middle classes’? And if their defining characteristic is ‘middle’, what lies above and below them?

In most discussion, whether in mainstream academia or in the media – ‘classes’ are seen as separate groupings, based on a shifting amalgam of occupation, income, status and lifestyle. Usually these classes are understood to be arranged in a hierarchy. Social mobility is often a major concern, and it is measured by looking at how many people move upwards (or indeed downwards) from one such designated class to another.

This approach creates many difficulties when tracking class affiliation through time. If we take a stripped-down approach in which occupation alone defines class, the most common traditional distinction is between blue-collar or manual workers, defined as ‘working-class’, and white-collar or non-manual workers, defined as ‘middle-class’. Given the powerful occupational shift in countries like the UK from manual to non-manual work, this yields the result, comforting for many commentators, that these  countries have in recent decades become more middle-class. 

Last year, a major academic survey of classes in the UK ended up with no fewer than seven ‘classes’. The survey, led by sociologist Mike Savage, sought to broaden the traditional occupational analysis of class by more fully taking into account ‘the role of cultural and social processes in generating class divisions’. The survey argues that ‘this new seven class model recognises both social polarisation in British society and class fragmentation in its middle layers’.[1]

The authors identify distinct groupings in terms of their experiences, attitudes and lifestyles, and relate them to underlying economic and social trends.  Such an approach is attractive because it roots class identity in something common to all of us, namely a life-path that can be mapped out and analysed, and because the data generated in such a survey can then be subjected to sophisticated statistical analysis. The elements selected for recording are underpinned by a particular conceptual framework, developed some thirty years ago by the French sociologist Pierre Bourdieu.[2]  In this approach, individuals are differentiated by their possession of economic, cultural and social capital in different quantities and proportions – the three forms of ‘capital’ being in principle independent of each other. The data are then analysed in order to identify clusters of individuals – eventually in this case, seven in number – who broadly share the same economic, social and cultural characteristics.

The immediate problem with this approach is that it avoids any need to identify or explain any overall social processes that might, between them, be shaping changes in how we ‘cluster’ in particular classes, and how these classes interact with each other. This makes it very hard to ask really important questions about society as a whole, and how social differentiation changes through time: in short, you can’t see the wood for the trees.

For this purpose, we could try a different approach. Specifically, it is helpful to go back to Marx; not the usual caricature of the ‘industrial proletariat’, but the actual concept of class in Marx. He argued that class was not a thing, but a relationship, that it was the key constitutive feature of any distinct form of social order, and that it was grounded in how a given society produced and reproduced its means of subsistence. In capitalism, the class relation is substantively what, for the purposes of critiquing political economy, he alternatively calls ‘capital’. This concept of capital is utterly different from that of Bourdieu: far from being an individual possession, it is the overall social relation between a class of people who own the means of production, and can live off the returns on those assets; and a class of people who are obliged instead to sell their labour-power in order to live and reproduce themselves.  Marx's familiar names for the two sides of this bilateral relationship are the bourgeoisie and the proletariat, but given the widespread unease today about such terminology we might just as well call them the ‘owners’ and the ‘workers’.

The great value of this relational approach to class is that it allows us to link firmly together those aspects of our society that are, in mainstream social sciences, treated separately: the economic, the political and the socio-cultural. Using this approach, we can immediately understand that occupational changes have no necessary impact on the proportions of these two ‘great classes’. We can also see that there are huge differences in income within each of them: among the class of owners between, for example, Bill Gates and the owner of the White Rose Hotel in my home village; or among the workers between a university vice-chancellor and a university cleaner.

A key element of the relational concept of class is to do with economic security. Anyone securely lodged within the owning class is thereby largely spared the ups and downs of the trade cycle, but small owners (Marx’s ‘petty-bourgeoisie’) may find themselves rudely relegated into the working class.  Some workers, usually among those termed ‘skilled’, are so essential to any continuity of production that they have permanent economic security, at least through their working lives. The issue of economic security (also known sometimes as ‘existential’ security, though that is also determined by natural disasters, war and peace, etc.) is of course historically the source of the introduction of measures of social protection and welfare. Because it is such an important matter, it is also one of the most important factors generating political parties and other representative institutions (notably of business and labour).

What, then, are the ‘middle classes’ in this approach? For this we can go way back before Marx, to the dawn of the industrial age. There we find a class of wealthy and powerful owners of land and businesses, and ordinary workers in agriculture, urban production and commerce, typically propertyless but retaining some traditional rights of access to means of subsistence. But between these two broad groups lay the ‘middling’ classes: yeoman farmers, independent artisans, urban craftsmen with employees, professionals such as lawyers and the clergy, and small owners in transport and trade.  With industrialisation, the occupational and spatial characteristics of these middle classes were thrown into disarray. The owners became more oriented to industry and finance, and less to land and commerce, while in some spheres, economies of scale in the new technologies rapidly differentiated them.  The new ‘industrial’ workers included both deskilled artisans and newly-urbanised ‘free’ labourers, as well as workers in transportation, commerce, government and household service; while even the still-numerous agricultural workers found their lives transformed by changes in technology and organisation. Crucially, the workers were differentiated right away both by increasing diversity of occupations, trades and skills, and by the owners’ need, in every economic sector, for a stratum of overseers and ‘white-collar’ clerks to record the flows of goods and money.

These changes can be tracked historically ever since; what is more, the social processes that were set in train in Britain as pioneer of capitalism can be seen to emerge as this new social order spreads. The phenomena of ‘middle class revolt’ that are seen today in Turkey or Brazil are strikingly similar, not just structurally but in terms of political agency, to Chartism in 1840s Britain. Then as now, workers who are more skilled, more educated and more urban are in fact responding more readily to the dynamics of the class relation between owners and workers. They are, for the most part, workers in Marx’s sense, whether they are journalists, car workers, teachers, fast-food sellers, students or whatever. They will find support, then as now, from enlightened owners who are shamed by their relative condition into seeking alongside them a more equitable distribution of wealth and power.  And they will seek political representation.

In identifying ‘classes’ as groups of citizens sharing common characteristics, studies like the recent LSE survey are undoubtedly useful, both for private sector firms in designing their marketing strategies, and for government departments and advisers on public policies. And part of their usefulness lies in identifying the extent to which individuals are able to move from one group to another – for example, from the LSE survey’s ‘established middle class’ (class 2) to the ‘elite’ (class 1), or from the ‘new affluent workers’ (class 4) to the ‘technical middle class’ (class 3). However, the relational approach to class offers a much more coherent account of the economic and social wellsprings of differentiation and dissent. And that is exactly why it is today largely passed over in silence by mainstream academia. 

But there is a further reason why the purveyors of knowledge should prefer instead to divide us into ‘working-class’ and ‘middle-class’, namely that they themselves have largely thrown in their lot with the owners.  Social scientists such as Alvin Gouldner, Lawrence King and Iván Szelényi[3] have argued that intellectuals – or as many would put it today, knowledge workers – sought in the 20th century to secure their relative position in society by playing upon the owners’ need for their scientific, technical and managerial skills. 

That worked reasonably well, as long as capitalism was contained more or less within national political boundaries, within which a consensus could be built around a welfare state, a mixed economy and a modicum of redistribution.  But in recent decades, the revival of liberalism, the rise of finance and global economic integration have shattered the compromises of welfare-state capitalism and Third World developmentalism alike. The owners can now roam the world in search of profit, recovering the share of wealth and power that they enjoyed in the late 19th century. They are now unchallenged by any alternative order, whether based on state socialism or on nationalism underpinned by ethnicity, religion or culture.

But they still require the knowledge workers. It is up to us (for I too am one, and so in all probability are you) to decide where our loyalties lie. Hopefully we will realise that, with the exception of the top class identified by the LSE survey – the elite – the rest of us may, despite all our apparent differences, together form a unified working class around our desire for a more equal and just society.

Hugo Radice is a Life Fellow at the School of Politics and International Studies at the University of Leeds.


[1] Mike Savage et al., ‘A new model of social class? Findings from the Great British class survey experiment’, Sociology 47/2, April 2013, p.  220.  At

[2] Pierre Bourdieu, Distinction.  London: Routledge, 1984.

[3] Alvin Gouldner, The Future of Intellectuals and the Rise of the New Class, New York: Seabury, 1979; Lawrence King and Iván Szelényi, Theories of the New Class: Intellectuals and Power, Minneapolis: University of Minnesota Press, 2004.

All comments are moderated, and should be respectful of other voices in the discussion. Comments may be edited or deleted at the moderator's discretion.

Remember my personal information

Notify me of follow-up comments?