An Absence of Fraternal Concurrence: A review of Dan Evans' A Nation of Shopkeepers

Perhaps one of the few things all socialists (however loosely one chooses to define the latter term) can agree on is that a tension exists between different social groups within capitalism. The content of that tension - and indeed of the social groups involved - is a subject that constantly rears its head, though is seldom given systematic treatment. Perhaps that isn’t surprising, particularly in the context of the eclipse of the 20th century workers’ movement. With the political primacy of class no longer a given, Dan Evans’ A Nation of Shopkeepers: The Unstoppable Rise of the Petty Bourgeoisie is a welcome attempt to grapple with the changing class composition of Britain. I have wanted to review the book for some time, and offer a somewhat different interpretation of the current class dynamics in Britain.

Class and the left

Although A Nation of Shopkeepers is primarily about the petty bourgeoisie’s role as a driving force of political change, Evans combines this analysis with a critique of the left’s conception of class, one that has lost sight of its structural basis:

“The excision of class and materialism and the relentless focus on individuals, coupled with the shift in the understanding of power, paved the way for the left to be subsumed wholesale not just by identity politics, but by the infantile liberal view of the world in which people are no longer moulded by structural forces beyond their control, but are now essentially innately good or bad, either reactionary or progressive - particularly racist or non-racist”.1

This is not an uncommon position. In the perpetual arguments that set class and identity at opposing poles, Evans comes out on the side of what is commonly called class reductionism, which itself is a slippery term. What do we mean by class reductionism? One class reductionist framework is that in which class is fundamentally determinant at the structural level (meaning that we can identify the objective structural antagonisms at the level of class - the point of production), whilst acknowledging that political interests may not map along class lines. The alternative framework is that in which abstract class categories are reified, resulting in the rejection of the complexity of social structures. Here I find the framework of McCarthy and Desan’s essay, The problems with class abstractionism2 useful to split these two conceptions of class reductionism apart, with the former definition referred to as class dynamism, and the latter as class abstractionism. I linger on this distinction because I have always labelled my own position as that of class reductionism, and by this I mean in the dynamic sense of McCarthy and Desan, and have found their essay useful for distinguishing between this dynamic class reductionism, and those class reductionists who give an a priori political primacy to class (class abstractionists). It’s worth noting that Evans does grasp this cleavage, and he exhibits a sharp understanding of the distinction between the structural underpinnings of class, and how class is articulated in reality (Evans frames this distinction in terms of objective and subjective class). What Evans accuses the left of is what McCarthy and Desan would call class relativism – where class loses its structural specificity and instead becomes a generic term for subjective group formation. However before we can address the content of his critique we must ask, who is the left?

Given much of the book is a critique of the British left, this is no minor question. It is puzzling then, that the left as a category remains undefined for the most part, whilst the lack of citations and evidence throughout leaves much of Evans’ analysis feeling like it is based more on a general vibe than any concrete data (I am told that the publisher Repeater Books, tends not to include citations in their publications). To be sure, “the left” is used loosely in common parlance, on the assumption that any audience on social media or in person is likely to understand what we mean. Although such an informal approach in an informal environment is understandable, it inevitably raises problems, evident in Evans’ own claim - notably in a book on class, not a casual chat - that the likes of current Labour Party leader Keir Starmer and previous Labour Party leadership candidate Owen Smith are “the left at the upper echelons”.3 When hardline stances on immigration and an economic platform that would sit just as well on the other side of the aisle of the House of Commons do not exclude one from the left, it would suggest that the category is so broad as to contain little meaning.

The final chapter gives a little clarity here, as Evans zones in on Corbynism as the object of critique. Evans contends that Corbynism,

“did not and could not engage with the working class precisely because it was animated by the desires, needs, interests and ideology of the new petty bourgeoisie and the professional managerial class; those involved simply did not understand or appreciate the scale of the glaring class divide between them and the people they wanted to appeal to.”4

Here we find the culmination of Evans’ analysis, one that starts by tracing the emergence of the new petty bourgeoisie, largely through the work of Nicos Poulantzas. If we take the left to mean Corbynism however, we are faced with the problem of incoherence. Given that Corbynism was centred around the Labour Party, drew in Trotskyist groupings as well as the likes of Open Labour, and necessarily entailed working with Labour Party MPs and members with politics hostile to that of Corbyn (for example, the aforementioned Owen Smith), one might describe Corbynism as analogous to a popular front, rather than anything more coherent. What does it mean to speak of Corbynism’s petty bourgeois composition? And more importantly, what effect would this petty bourgeois outlook have, considering most Labour Party and/or Momentum members had little capacity to affect the direction Corbynism took, beyond turning up to Labour Party meetings and knocking on doors? Even if we take the class composition outlined by Evans as given, the causal links between the ideological outlook of mostly powerless individuals and the direction of a highly centralised parliamentary campaign are missing from his analysis.

It would however be a mistake to take Evans’ diagnosis of the class composition of the left as given. It is important to ask whether he is right, and whether the prevalence of the new petty bourgeoisie has been central to the left’s recent failings. To answer this question, we can turn to his conception of the new petty bourgeoisie.

The new petty bourgeoisie

The outsourcing of manufacturing from the Global North to the Global South, and the shift towards service driven economies in the former has been well documented, particularly in the case of Britain, where the loss of colonial holdings saw the fall of a global power, and the adaptation to a new role as the hub of global finance (documented in Tony Norfield’s tour de force, The City5). The decline in traditional industrial working class jobs saw a shift to service sector employment. These white-collar workers are Evans’ new petty bourgeoisie. Like the proletariat (and unlike the old petty bourgeoisie), they do not own means of production, but for reasons that are social, cultural, and ideological rather than economic, they are not of the proletariat. The commonality between the old petty bourgeoisie (petty proprietors owning their means of production) and the new petty bourgeoisie is in their fluid class position. In seeking expansion into the capitalist class whilst simultaneously fearing falling into the same plight as workers if their small business fails, the old petty bourgeoisie have traditionally occupied a fluid class position that engenders individualism and self-interest. Although the new petty bourgeoisie do not own their means of production, they self-identify as middle class and fixate on social mobility. Higher education and habitus are important factors for Evans here, with mental labour as a dividing line. The new petty bourgeoisie’s relationship to education has not remained static, however. Where once it was the mechanism by which to bootstrap oneself into the middle classes, it now lends to a social anxiety born of the devaluation of a university education which no longer guarantees well paid secure jobs. By contrast, Evans plays down the role of home ownership in class determination. Whilst acknowledging the vast increase in owner-occupation towards the end of the 20th century and the start of the 21st, and that this change has had complicating effects on the interests of working class homeowners, Evans (correctly in my view) rejects any framework that determines class purely based on housing ownership because most owner occupiers do not profit from home ownership.6

How does one emphasise the cultural and ideological effects of education, whilst playing down the economic effects of home ownership on class determination? Evans turns once more to Poulantzas, and the concept of class instinct. Working class instinct, it is claimed, is anti-hierarchical and anti-authoritarian. The worker knows which side of the class divide they are on within the despotism of the workplace. By working in a collective environment and possessing this class instinct, the working class do not internalise or reproduce the relations of domination found in the workplace. By contrast, the new petty bourgeoisie, represented by police, doctors, civil servants, teachers, engineers etc are all part of the repressive and ideological state apparatus. The “ideology of the agents is mediated and formed by the functions it carries out.”7 Their class position is located in - and legitimated by - hierarchy. The new petty bourgeoisie then, are defined by isolation and individualism, whilst the working class are defined by collectivism and hostility to individualism. Whilst home ownership might complicate class position, its effects on individual interests are less significant in terms of class determination than the legitimising effects of being located within the ideological state apparatus.

A stratum or a class?

Evans’ narrative is a compelling and intuitive one, but there are numerous inconsistencies. A focus of his narrative is the individualistic nature of the new petty bourgeoisie, and yet he also notes that – as a result of degrading workplace conditions - they are more highly unionised than the working class. A problem with this formulation is that the terms new petty bourgeoisie, professional managerial class, middle class, and white collar workers are used interchangeably, as are class and strata. Do the police and graduate call centre workers have the same class interests, and are both part of the state apparatus? Probably not. This is not necessarily a problem, if the focus of Evans’ analysis is on a class stratum (as is the case for Erik Olin Wright, or Poulantzas’ use of social categories). The difference here is important. A stratum can cross class lines and enables us to identify fractional interests within and between classes, without losing the specificity of a class’s relation to production. What are the consequences of conflating strata with class? We can turn to Simon Clarke’s critique of Poulantzas’ theory of the state. For Clarke, Poulantzas separates the political and ideological from relations of production:

“The economic level is that of material production, guaranteeing the physical survival of the whole. The political level assigns individual agents to means of production as owners or non-owners, the latter being residually owners of labour-power, and so as recipients of their respective revenues. The ideological level constitutes these individual ‘supports’ of the relations of distribution as social subjects able to fulfil their roles in society. The economic level is thus the technical realm of material production, the political and ideological levels are the social realm which establishes the social conditions of material reproduction. For this analysis, therefore, the autonomy of the political and ideological relative to the economic is the supposed autonomy of relations of distribution relative to relations of production which depends on the bourgeois conception of production. The consequence is the view of social relations as constituted not in production, but ‘politically and ideologically’, or ‘normatively’, which in turn underpins a reformist politics.”8

There’s a lot going on here, so it is worth breaking it down. Clarke shows that by splitting the political and ideological realms off from the economic, the former are given primacy over the latter. In simple terms, the dominance of politics and ideology results in interest groups conflicting over a share of the social product (referred to by Clarke as the bourgeois conception of production). This social product is treated as fixed because the struggle is no longer at the level of the economic (production), but at the political and ideological (ie distribution of the fixed product). This inevitably leads to reformism because interest groups are struggling over a larger slice of the product, as opposed to control of production itself.

How does Evans do this? The new petty bourgeoisie – despite sharing the proletarian relation to production – are classified with the old petty bourgeoisie on ideological and political grounds, despite having a different relationship to production to the latter. It is also why Evans can claim that the dominance by the professional-managerial classes and the new petty bourgeoisie of the labour movement is a direct reason for the latter being “pathetic” and the old petty bourgeoisie supporting big capital. This is an extraordinary claim to make in passing, one that fails to account for the long decline of the global workers’ movement (a movement that “destroyed itself as a force for change because it aimed at preserving the proletarian condition, not superseding it”). Similarly, the left’s roots in the state bureaucracy came before deindustrialisation, with the workers’ movement putting its faith in the development of the productive forces, a process well documented in EndNotes410 and Robert Michels’ Political Parties11 (Michels who infamously rejected social democracy for fascism precisely because of this tendency towards bureaucratisation – his so-called iron law of oligarchy – several decades before deindustrialisation).

Returning to the conflation of strata and class, I do think that Evans identifies a real phenomenon in the sense of divergent interests of the PMC. However, the utility of the PMC concept is in identifying a cross-class stratum that exists in both the working class and the petty bourgeoisie. This means going against the Ehrenreichs’ own formulation of the PMC as “a class totally distinct from the petit bourgeoisie.”12 The cultural and ideological differences that Evans highlights are ofcourse real, but it is the relation to production that must inform our structural class categories if we are to maintain a consistent revolutionary politics, one that goes beyond centring relations of distribution. With that in mind we know that many in this stratum (most white collar workers that are not in management positions) share the same relationship to production as the proletariat, whilst those in management positions sit as an intermediate layer between labour and capital (as do the traditional petty bourgeoisie), living with the class mobility/precarity that comes with it. Evans claims that class mobility is the mechanism that stitches together the old and new petty bourgeoisie, whilst the working class is a static class. This misses the process of lumpenisation, to which we now turn.


According to Evans, the working class are a static class, and because they are static, “they don’t experience the fear of falling that the petty bourgeoisie goes, because there is nowhere to go.” This ignores what I would argue is the dominant class dynamic at present – that of lumpenisation. In volume 1 of Capital Karl Marx describes the process whereby capital accumulation initially absorbs society’s excess labour. This initial process of accumulation, whereby the labour process has been subsumed under capital, but not yet revolutionised by it, is termed formal subsumption of labour.13 As accumulation proceeds however, centralisation occurs as larger firms outcompete and absorb smaller firms via the productivity advantage given by newer technology, which in turn requires less labour to produce more commodities. This is the revolutionising of the labour process, known as real subsumption of labour.14 These processes are not linear however.  Formal subsumption in a sector results in an absorption of society's excess labour, but real subsumption via a rising organic composition of capital reverses this, contributing to an increase in the relative surplus population. This excess population might then be absorbed by another sector in which the accumulation process is still in its infancy. These two moments within the accumulation process describe how the relative surplus population is absorbed and sloughed off.

What happens when there is no profitable sector left for capital accumulation? The answer is right before our eyes. As Aaron Benanav argues in Automation and the future of work,

“global capitalism is failing to provide jobs for many of the people who need them. There has been, in other words, a persistently low demand for labor, one which is no longer accurately registered in unemployment statistics. Labor underdemand is reflected in higher spikes of unemployment during recessions, as in the 2020 pandemic recession, and in increasingly jobless recoveries, a phenomenon likely to be repeated in the pandemic recession’s aftermath.”15

What are the consequences of this increasing underemployment for class dynamics? Here I want to turn to Clyde Barrow’s The Dangerous Class, a treatment of the lumpenproletariat that I have found both immensely useful and relevant to the current moment. Like Benanav, Barrow identifies how rising structural unemployment and underemployment are a result of reduced demand for labour power, generating casual, low-wage and informal labour markets. By excavating Marx’s treatment of the lumpenproletariat, Barrow shows how the distinguishing feature of the lumpenproletariat is its nonrelation to production:

“The lumpenproletariat’s nonrelation to production means that it is not structurally organized by capitalist relations of production, and therefore it cannot develop an independent class consciousness or any sense of a historical mission within capitalism.”16

Whilst this much is generally accepted, Barrow argues that the category goes beyond simply the dispossessed (ie those who cannot or will not work), including also those who are casually, seasonally, or extremely irregularly employed. Like Evans, Barrow draws on Poulantzas’ social categories, though he is careful to emphasise that the lumpenproletariat are not a class but a fluid stratum, constantly undergoing a process of dynamic recomposition as elements of the relative surplus population are sloughed off into it or removed based on the decomposition and recomposition of the working class.

Coming back to A Nation of Shopkeepers, the working class has clearly never been a static category, and in light of increasing underemployment it is clear that its decomposition is a dominant dynamic as capital accumulation stalls. According to the ONS, the number of regular active workers in the UK gig economy more than tripled between 2016 and 2022, constituting 22.1% of the total workforce. Although by no means the focus of his book, Evans does describe platform workers as occupying a grey area between the working class and the petty bourgeoisie. Instead, Barrow draw us to Marx’s words on the stagnant surplus population (which Barrow includes as part of the lumpen category):

“a part of the active labour army, but with extremely irregular employment. Hence it offers capital an inexhaustible reservoir of disposable labour-power. Its conditions of life sink below the average normal level of the working class, and it is precisely this which makes it a broad foundation for special branches of capitalist exploitation. It is characterized by a maximum of working time and a minimum of wages.”17

Marx is characteristically prescient here. A recent study by the University of Bristol found that over half of gig economy workers are earning below minimum wage. These low wages are often used to supplement income from other employment. I think it is clear that the lumpen category is appropriate here, rather than quibbling over whether Uber drivers do or do not own their own means of production (a debate which Evans correctly sidesteps). Evans is also correct in identifying the atomisation of gig work, and its effects in placing workers in competition with one another (though I feel he ignores that this is also the case within the working class). These dynamics have implications for political subjectivity, discussed in the next section. Crucially, the analysis presented here (which rests heavily on Clyde Barrow’s work) enables us to understand the changing class dynamics in the UK at the level of production, without losing sight of the ideological effects of this process of lumpenisation.

The question of working class identity

Returning to Corbynism, Evans explains part of its failure in the movement’s inability to appeal to working class interests and a preoccupation with social justice issues. In many ways this comes back to the tired debate between class reductionism and identity politics. McCarthy and Desan are once again useful here. On the face of it, Evans is rejecting class relativism, but as I hope I have convincingly argued, this is the very trap that Evans falls into. By giving primacy to the political and ideological levels over the economic, “class loses its economic specificity and instead becomes a generic term for subjective group formation.” By peeling the new petty bourgeoisie off from the working class, Evans is able to directly target where the problem lies.

Yet whilst class relativism dominates his account of the new petty bourgeoisie, when it comes to the working class we find Evans’ account is class abstractionist. For Evans, the working class’ structural relation to production means that they are his a priori political subject. I would argue that the reality is a great deal messier, because it involves grappling with the atomised nature of the working class, and the fractional group interests that are at play within said class. Though this fractionation of interests is more strongly the case today than it was previously, it was always the case! In A History of Separation the EndNotes collective identify the internal limit on the workers’ movement as that set by the limits of working class identity:

“only a portion of the proletariat ever identified with the programme of the workers' movement. That was because many proletarians affirmed their non-class identities - organised primarily around race and nation, but secondarily around gender, skill and trade – above their class identity. They saw their interests as adding up differently, depending on which identity they favoured.”18

Whilst the proletariat’s revolutionary potential is set by their relation to production, within that structure are non-class identities which often appear thicker and more binding for the in-groups concerned. This shouldn’t come as a surprise when these groups are constantly pitted against one another. By untethering the new petty bourgeoisie from the working class as Evans does, things become much simpler as we can point to our revolutionary and reformist subjects. The alternative is a much muddier conception of a working class with disparate and often incoherent interests that – whilst united by their relation to production – have been made diffuse by both economic and ideological/political forces. Again, from EndNotes:

“The time of the workers' movement was simply the time of the rise and decline of the semi-skilled male worker and of the industries where he worked. Together they made it possible to imagine that capital was tendentially unifying the class by means of an affirmable workers' identity. But it was only insofar as those industries were expanding that the workers' movement could see the semi-skilled worker as its future being realised in the present. Once those industries went into decline, the glorious future declined as well.”19

What this suggests is that we need to stop simply overlaying the old models of the workers’ movement onto new terrain. The nature of work, and of capital accumulation have considerably changed since the apex of the workers’ movement. As such we cannot simply expect that a new workers’ movement built in the image of the previous one will be any more revolutionary or successful. The revolutionary potential of the proletariat remains, though the revolutionary subject remains as illusive as ever.

Where next?

In many respects the processes of lumpenisation and underemployment were already being theorised decades ago by the Black Panthers. Eldridge Cleaver saw the lumpenproletariat as including anyone who does not have a direct economic relation to production. Quoting from Clyde Barrow again,

“Thus, instead of spending so much effort trying to identify a new working class, or to redefine the nonindustrial composition of the working class, iit was necessary to recognize that the fundamental process of class composition and decomposition in contemporary capitalism was the lumpenization of humanity across the world based on humanity’s physical nonrelation to the means of production. Based on this definition, by 1972, the lumpenproletariat, in one form or another, was the majority of the population in nearly every capitalist country on earth.”20

It is clear therefore, that lumpen consciousness must be part of any revolutionary transformation. Because the lumpen are excluded from production, they exist in a state of dependency. Concurrently, the lack of worker power (as evidenced by the decades-long downward trend in trade union membership - the recent strikes across multiple sectors in the UK are encouraging, though they are defencist in nature, attempting to prevent real term pay cuts) has further exacerbated dependence on the state to contain inequality, whilst at the same time the welfare state is stripped further and further back. This leads us to a contradictory movement in the role of the capitalist state because although the welfare state acts as a weapon of dependency, the shedding of the welfare state is proving necessary to optimise accumulation, setting hard limits on any reformist agenda.

Although understanding these processes does not necessarily highlight a clear path forward, we are able to seriously think about which models we can reject. The appeal of class reductionism in the abstractionist sense of McCarthy and Desan is that it provides us with a ready-made revolutionary subject, but this ignores the incredibly messy history and modern day of class composition. One of the preoccupations of The Black Lamp has been to reject old models that are inappropriate for the current moment. The class and accumulation dynamics in the UK (and across much of the globe) point to a need to thoroughly understand the ongoing process of lumpenisation. Whilst I’ve been critical of A Nation of Shopkeepers, I found the book useful in helping me think through these processes, and I applaud Evans’ desire to move beyond a simple worker-capitalist dichotomy and towards a more thoroughgoing treatment of class in the UK. The point is ofcourse to change the world, but we need a much better understanding of it to really think about how to do that.

Works Cited

  1. Dan Evans, A Nation of Shopkeepers: The Unstoppable Rise of the Petty Bourgeoisie (Repeater Books, 2023), p14.
  2. Michael A. McCarthy & Mathieu H. Desan, The problem of class abstractionism (Sociological Theory, 2023, Vol. 41(1) 3–26).
  3. Dan Evans, A Nation of Shopkeepers: The Unstoppable Rise of the Petty Bourgeoisie (Repeater Books, 2023), p30.
  4. Ibid, p.276.
  5. Tony Norfield, The City: London and the Global Power of Finance (Verso, 2016).
  6. Dan Evans, A Nation of Shopkeepers: The Unstoppable Rise of the Petty Bourgeoisie (Repeater Books, 2023), p246.
  7. Ibid, p.167.
  8. Simon Clarke, Marxism, Sociology and Poulantzas’ Theory of the State (Capital & Class, 1977, 1(2), 1–31).
  9. Gilles Dauve, “Foreworld: Out of the Future” in Eclipse and re-emergence of the communist movement (PM Press, 2015). p.27.
  10. EndNotes Collective, “A History of Separation” in EndNotes4: Unity in Separation (EndNotes Collective, 2015).
  11. Robert Michels, Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (Martino Fine Books, 2016).
  12. Barbara Ehrenreich & John Ehrenreich, “The new left and the professional-managerial class” in Radical America (Alternative Education Project, 1977, 11(3), 7-24).
  13. Karl Marx, Capital Volume 1 (Penguin Classics, 1990). p.1019.
  14. Ibid, p.1023.
  15. Aaron Benanav, Automation and the future of work (Verso, 2020). p.9.
  16. Clyde W. Barrow, The Dangerous Class: The Concept of the Lumpenproletariat (University of Michigan Press, 2020). p.15.
  17. Karl Marx, Capital Volume 1 (Penguin Classics, 1990). p.795.
  18. EndNotes Collective, “A History of Separation” in EndNotes4: Unity in Separation (EndNotes Collective, 2015). p.127.
  19. Ibid, p.108.
  20. Clyde W. Barrow, The Dangerous Class: The Concept of the Lumpenproletariat (University of Michigan Press, 2020). p.107.

Vivak Soni is an editor at The Black Lamp.