Marx and Weber Compared as Working Schemes

Part 4 covered Marx and Weber as thinkers. This lesson treats their accounts of class as what they also are: rival instructions for sorting a real population into groups.

That is a more demanding test than it sounds. A scheme has to say what to do with every case, produce groups that behave differently, and explain why the line falls where it does. Run the two schemes over the same five people and they disagree — not everywhere, but at exactly the points where most of the modern population sits.

Five people, two schemes, and the places they disagree.

Ravi owns a small engineering firm. Twelve employees, two machines he bought outright, a workshop he owns.

Sunita is a nurse on a hospital ward, sixteen years qualified, employed on a national pay scale.

Anil is a self-employed builder. He owns his tools and his van, works alone, occasionally takes on a labourer for a fortnight.

Meena runs a shop she inherited. Two part-time staff. She works behind the counter herself, six days a week, and earns less than Sunita.

And Farooq is a senior civil servant. He owns no productive property whatsoever. He controls a budget of two hundred million and a staff of four hundred.

Run Marx's scheme.

The question is relation to the means of production: do you own productive property and buy others' labour, or do you sell your own?

Ravi is bourgeoisie — clearly, and the scheme is unambiguous. Sunita is proletarian. She owns no productive property and sells her labour. That her work is skilled, licensed and socially valued changes nothing about the relation. Anil is petty bourgeoisie — owns his means of production, employs essentially nobody, works with his own hands. A category Marx recognised and expected to be squeezed out of existence. Meena is also petty bourgeoisie, though she earns less than the nurse the scheme calls proletarian. And Farooq is the problem. He owns nothing and sells his labour, so the scheme says proletarian. He also commands four hundred people and a fortune in public money. Something has gone wrong.

Now run Weber's.

The question is market situation: what can you bring to market, and what life chances does it buy you?

Ravi has property — the primary division, and the one Weber shares with Marx. Sunita has a scarce, credentialed, licensed skill. Her market situation is not the same as an unqualified worker's, because entry to her occupation is restricted and her qualification is portable. The scheme separates them; Marx's does not. Anil has property but no organisation, and his market position depends on demand for a service he supplies personally — a distinct situation again. Meena has property that generates a low return. Weber can say this without embarrassment: she is propertied and her market situation is poor. And Farooq is handled without difficulty, because Weber has two further dimensions. His status — the honour attached to his position, the deference he receives, the circles he moves in — is high. His party position, in the sense of organised power to pursue interests, is very high indeed. His class situation in the narrow sense is that of a well-credentialed employee; his position in the structure of advantage is not captured by that alone.

Where the two schemes agree, they agree strongly. Ravi is at the top on both. Both treat class as a position rather than a feeling.

Where they disagree is Sunita, Meena and Farooq — a nurse, a shopkeeper and an official. That is not a peripheral set of awkward cases. That is most of a modern population.

A class scheme is a claim about where the important lines fall — and it is testable.

Every scheme partitions a population. The question that decides between them is not which is theoretically deeper but which cuts the population where outcomes actually differ.

If a scheme is right, then the groups it produces should differ from one another — in life expectancy, in children's chances, in security, in health, in what happens to them in a recession — more than they differ internally.

That is a checkable claim (see 7.9.1), and the answer varies by what you are trying to explain. A scheme that best predicts mortality is not necessarily the one that best predicts political behaviour, and knowing which question you are asking determines which scheme you should reach for.

Which is the useful way to hold the Marx–Weber argument. Not as a contest between two worldviews to be won, but as two instruments with different measurement properties (see 7.2.2).

The two schemes, precisely

Marx: class is a relation, and the relation is exploitation.

The axis is ownership of the means of production, and everything follows from it.

Those who own productive property can live from it. Those who do not must sell their capacity to work. And in the production process, the second group produces more value than they receive back in wages — the surplus is appropriated by the first (see 4.1.3). This is what makes the relation exploitative in a technical rather than a moral sense: one group's position depends on the other's, and the gain of the one comes from the labour of the other.

Three consequences follow that distinguish this scheme from every other.

Class is relational, not gradational. Classes are not rungs on a ladder of income. They are defined by their relation to each other, and you cannot describe one without the other — which is why a scheme built on income bands is not a class scheme at all in this sense.

Class positions carry structurally opposed interests. Not opinions — interests. Whether people recognise them is a separate question, and the gap between the two is the whole subject of class-in-itself and class-for-itself (see 4.1.8).

And the scheme predicts polarisation. Intermediate positions get squeezed as capital concentrates; the petty bourgeoisie shrinks; society tends towards two camps.

On that last prediction the record is mixed and the honest verdict must be stated. Property ownership did concentrate. But the intermediate positions did not disappear — they multiplied, into a vast array of managerial, professional, technical and administrative occupations that own nothing and are plainly not in the position of a factory labourer. Farooq is not a theoretical curiosity. He is one of tens of millions, and accounting for him has been the central task of Marxist class analysis ever since.

Weber: class is market situation, and it is one of three dimensions.

Weber's definition is narrower than people expect and it is precise: a class situation is a market situation — the typical chances a person has of obtaining goods, living conditions and personal life experiences, given what they can bring to a market.

The primary division is property and its absence, which he takes over from Marx directly. The second division is what he adds: among the propertyless, market situation varies enormously by what services they can offer and how scarce those are. The nurse and the labourer stand in a different relation to the labour market, and Weber's scheme registers that as a class difference rather than a variation within a class.

And class is one of three dimensions of the distribution of power (see 4.3.6).

Class — the economic order; market situation; what your resources buy. Status — the social order; honour, prestige, lifestyle, whom you may marry and eat with, whose company confers standing. Party — the political order; organised association to pursue interests and influence.

They are related and do not coincide. The impoverished aristocrat has status without class. The wealthy outsider has class without status. The trade union official has party position without either. And Farooq has moderate class, high status and high party — a combination Weber's scheme describes precisely and Marx's has no vocabulary for.

Weber's mechanism for how class advantage is maintained is social closure: groups restrict access to opportunities by limiting eligibility to those with some characteristic — a credential, a licence, a membership, a name, a nationality. Exclusion from above and usurpation from below. This is 8.1.1's opportunity hoarding, and it does not require exploitation at all.

Two further Weberian points that matter.

A class is a category, not a group. People in the same market situation may or may not become a collectivity that acts. Weber thought status groups were much more likely to become real communities than classes were — because status involves shared lifestyle, sociability and mutual recognition, while class membership can be entirely invisible to its holders.

And there is no necessary polarisation. Market situations are plural, and there is no mechanism driving them towards two.

One — dimensions. One (ownership) against three (class, status, party).

Two — the mechanism. Exploitation: a relation in which one party's advantage comes from the other's labour. Market advantage: a distributional position in which one party can command more, without necessarily taking it from anyone.

Three — the shape. A tendency towards two poles, against an open plurality of market situations.

Four — collective agency. For Marx, classes have objective interests and the central question is whether they come to act on them. For Weber, whether a class becomes a group is contingent, and usually it does not.

Five — where skill sits. Marx's axis is ownership, so credentialed employees are proletarian. Weber's axis includes what you can bring to market, so credentials constitute a distinct class situation. This is the difference that decides what to do with most of a modern workforce.

What each does with the awkward cases

The manager problem, and the two best solutions.

Farooq — and every senior manager, administrator and professional who owns nothing and commands a great deal — is the case that forced both traditions to develop.

Erik Olin Wright's solution, from within Marxism, is the most rigorous attempt. He argued that such positions occupy contradictory class locations: they are exploited in relation to capital, and they simultaneously dominate and exploit others. Rather than forcing them to one side, the scheme recognises the contradiction as real.

He then generalised it. There are, he argued, three kinds of asset whose control confers advantage: capital assets (ownership of the means of production), organisation assets (control over the coordination of labour — what a manager has), and skill or credential assets (what a professional has). A twelve-cell scheme results, and the crucial move is that the second and third are treated as bases of exploitation in their own right rather than as mere market advantages.

Notice how far that travels towards Weber, and Wright acknowledged it — his later work engaged directly with closure theory and with the question of whether Marxist and Weberian class analysis are rivals or complements.

John Goldthorpe's solution, from the Weberian side, is the one that won institutionally.

His axis is the employment relationship, and it is beautifully simple. Employers face two problems with employees: how hard is the work to monitor, and how specific is the human capital involved? Where work is easy to monitor and skills are general, the employer uses a labour contract: payment for effort or time, easily terminated, no career structure. Where work is difficult to monitor and requires firm-specific or professional expertise, the employer must instead secure commitment — a service relationship: salary rather than wage, incremental progression, pension, job security, prospective rewards. The employer buys loyalty because supervision will not work.

This is why professionals and managers are a distinct class in his scheme, and it is a mechanism rather than a description. It generates the standard European class schema and, through it, the official classification used in British statistics (see 8.2.2).

And it handles all five of the story's cases: Ravi as employer, Farooq and Sunita in service relationships of differing degree, Anil and Meena as self-employed with distinct positions, and a labourer under a labour contract.

Three things "class" means that are constantly confused.

Class as position — a location in the economic structure, defined by ownership, market situation or employment relationship. What both schemes above are about, and what predicts outcomes.

Class as identity — what people say they are. These are only loosely related. Across many countries most people, asked, place themselves in the middle regardless of their position (see 8.1.1) — so a survey question about class identity and an occupational classification are measuring different things and will produce different results.

Class as culture — taste, accent, manners, consumption, comportment. Bourdieu's territory (see 5.8.1), and genuinely important, because cultural markers are how closure is operated in practice. But it is not the same object as position, and treating an accent as a class measure produces a classification of self-presentation.

And the commonest error of all: treating income bands as classes. Income is largely an outcome of class position, not the position itself. A scheme built on income cannot explain income, and it puts the shopkeeper and the nurse in the same box while separating the nurse from her colleague on the next pay point. Position is about the source and security of income, not its amount.

Because which scheme you use determines what you can find, and both are still in use.

On the empirical test, the Weberian employment-relations schemes have performed well at exactly the job class schemes are asked to do: they cut the population where outcomes differ. Health, mortality, children's educational transitions, security of employment, and what happens to people in a downturn all vary systematically and steeply across those categories, and they do so after controlling for income — which means the classification is capturing something income does not.

And the Marxist axis has kept doing work the other cannot. It explains why capital income and labour income behave differently, why ownership compounds across generations in a way earnings do not (see 8.3), why the return to capital is a distinct political question, and why some conflicts are structurally opposed rather than merely competitive. The concentration of wealth documented in 8.1.2 is a fact about ownership, and a scheme with no concept of ownership will not see it.

Three questions for any use of the word.

Position, identity or culture? These are three objects and most arguments about class fuse them.

Which axis — ownership, market situation, or employment relationship? And what does that axis do with a nurse, a shopkeeper and a senior official, who are where most of the population lives?

And is the scheme earning its place? Does it produce groups that differ more between than within, on the outcome in question? If not, it is a vocabulary rather than an analysis (see 7.9.1).

And one closing observation about the two founders, which is worth carrying. Marx's chapter on classes in the third volume of Capital — the place where he was finally going to define the concept — breaks off after about a page. The manuscript stops mid-argument. The most influential theory of class in history has no completed definition of its central term, which is either a scandal or a fair description of the difficulty, and probably both.

Five people expose where the two schemes diverge: an employer, a nurse, a self-employed builder, a shopkeeper and a senior official who owns nothing and commands four hundred people. They agree on the employer and disagree about the rest — which is most of a modern population.

Marx's axis is ownership of the means of production, and it makes class relational (classes defined by each other, not rungs on a ladder), gives positions structurally opposed interests regardless of awareness, and predicts polarisation. The last prediction is where the record is against him: intermediate positions multiplied rather than disappearing.

Weber's axis is market situation — what you can bring to market — which makes credentialed skill a class difference rather than a variation within a class. And class is one of three dimensions: class, status and party, which are related and do not coincide. His mechanism is social closure, which requires no exploitation. A class is a category, not a group, and status groups become communities more readily than classes do.

Five differences: one dimension against three; exploitation against market advantage; polarisation against plurality; necessary interests against contingent groupness; and where skill sits — the difference that decides what to do with most of a workforce.

Two solutions to the manager problem. Wright's contradictory class locations, generalised into control over capital, organisation and skill assets — which travels a long way towards Weber. And Goldthorpe's employment relationship: where work is hard to monitor and expertise is specific, employers must buy commitment with a service relationship rather than a labour contract. This is a mechanism, it handles every case, and it became the official classification.

Three meanings of class are constantly fused: position, identity and culture. And income bands are not classes — income is an outcome of position, and a scheme built on it cannot explain itself.

Empirically the employment-relations schemes cut the population where health, mortality, security and children's chances differ, net of income. And the ownership axis keeps doing what the other cannot: explaining the concentration and intergenerational compounding of wealth.

Marx's chapter defining class breaks off after a page.

Means of production — the property from which productive activity is organised.

Relational vs gradational class — classes defined by their relation to each other, versus rungs on a scale.

Exploitation — appropriation of surplus produced by another's labour; a relation, not an insult.

Class-in-itself / class-for-itself — an objective position; a group conscious of and acting on it.

Market situation — Weber: the life chances your marketable resources command.

Class, status, party — the economic, social and political orders of the distribution of power.

Social closure — restricting access to opportunities by eligibility criteria; exclusion and usurpation.

Contradictory class location — Wright: a position simultaneously exploited and exploiting.

Capital, organisation and skill assets — Wright's three bases of advantage.

Labour contract / service relationship — Goldthorpe: payment for effort under supervision, against salary, progression and security purchased where monitoring fails.

Class as position / identity / culture — the economic location, the self-description, and the taste and comportment that operate closure.

One — classify five people you know. Run both schemes over them. Note precisely where the two disagree — it will be the credentialed employees and the small proprietors.

Two — find the service relationship. Take any job you know well and ask: how easily is the work monitored, and how specific is the expertise? Then check whether the pay, security and progression match what Goldthorpe's mechanism predicts.

Three — separate the three meanings. Take a public argument about class and mark which sentences are about position, which about identity and which about culture. Most arguments switch between them without noticing.

Four — test a scheme. Find any statistic broken down by social class in your country and ask what classification was used. Then ask whether it is based on ownership, market situation or employment relations.

Five — do the Farooq problem yourself. Take a senior public official who owns no productive property and write down, in one paragraph each, what Marx's scheme and Weber's scheme say about their position. Then decide which description you would rather have if you were trying to predict their politics.