Chris Harman


The Working Class
After the Recession

(Autumn 1986)

First published in International Socialism 2:33, Autumn 1986.
Reprinted in A. Callinicos/C. Harman, The Changing Working Class, London 1987.
Copied with thanks from the International Socialism Website.
Marked up by Einde O’Callaghan for the Marxists’ Internet Archive.

ONE OF THE MYTHS feeding the policies of Labour Party leader Neil Kinnock and his supporters has been that of the ‘decline’ of the working class, or at least of the organised working class. It is an old myth, going back at least as far as a pamphlet Must Labour Lose? by the pollster Mark Abrams in the aftermath of the 1959 general election. It was revived in the aftermath of the 1983 election with Andre Gorz’s Farewell to the Working Class and the publication of various articles by Eric Hobsbawm in Marxism Today. After taking a bit of a nose dive during the miners’ epic battle the thesis is now with us again, with a keynote speech at the TUC by General and Municipal Workers Union general secretary John Edmonds harping on about ‘a new servant class’ which the unions could not expect to organise [1], and the labour editor of die Financial Times producing a much heralded book, Strike Free: New Industrial Relations in Britain. [2]

Gorz put forward the straightforward analysis that changing technology meant the working class was finished. Hobsbawm was usually more circumspect, even admitting on occasion that the majority of people were still ‘employed for wages/salaries’. [3] But his most popular argument for Labour politicians was that which held that ‘the manual working-class core of the traditional socialist labour parties is contracting and not expanding ...’ and that this could only be countered by ‘alliances’ with the middle class. [4]

Edmonds’ arguments are backed up by academics and Euro-communists who claim that trends in the economy and the workforce are splitting the working class in two, between a relatively small ‘core’ of organised worker trade unionists, and a ‘periphery’ of part-time and temporary, mainly women workers. [5]

The implication of all such arguments is that there is little future for class-based militancy. The arguments are wrong. Real changes have been taking place in the working class, and it is very important that socialists understand them. But they are rather different to those portrayed by Gorz, Hobsbawm or their admirers.

The manual industrial working class

The first, most obvious feature of the working class in Britain has been the decline in the number of workers in what the Department of Employment calls ‘production industries’. [6] Total employment in these industries (which included their white collar employees and foremen) fell by 22.8 per cent from 1978 to 1985, so that at the end of 1985 there were 5,928,000 employees in ‘production industries’, of whom 5,355,000 were in manufacturing. [7] This fall followed a previous decline in manufacturing jobs from 8,600,000 jobs in 1966 to 7,300,000 in 1977.

The significance of these changes can be seen by comparing them with what happened through the first sixty years of this century, when the numbers employed in manufacturing, especially in engineering and motors, grew dramatically. Employment in metal and engineering industries – these are rough figures, because of the way the Department of Employment has changed the categories into which it divides industries over the years; but they give an idea of the trend – changed as follows (figures in thousands):



. . . .




. . . .



. . . .



. . . .



. . . .



. . . .



. . . .



. . . .



. . . .



. . . .


It is the figures for the past 20 years which seem to give immediate plausibility to the thesis about industrial workers losing their importance. But a reduction in the workforce of an industry is not the same thing as the contraction of an industry. Redundancies and closures may occur in three different ways: as part of a process of slashing output, as a consequence of making existing employees work harder in a situation of stagnant or only slowly rising output, or as a result of capital investment which increases productivity faster than output. [8] Only the first of these involves deindustrialisation – the disappearance or the movement abroad of whole industries. The other two involve a continuing or even increased level of output, but with a smaller workforce.

The distinction is important. For although workers clearly lose their objective ability to exercise power against their employers after deindustrialisation they do not do so in the other two cases. A smaller workforce can be as potentially powerful as a larger one. As Batstone and Gourlay note in an important study of present-day trade union organisation, after cuts in the labour force:

‘The advantage to employers may be only temporary. Once the rate of job loss has receded, then union power is likely to return: for the ultimate base of union power within the workplace is not its absolute size, but its ability to stop production when the employer wants it’. [9]

The figures for industrial output show that deindustrialisation has been the exception rather than the norm [10]:







Industrial production





Manufacturing production





Output has risen in production industries as a whole, even if it has declined about 9 per cent in manufacturing. And within this there has been a massive rise in productivity [11]:






Manufacturing output per head





There are fewer industrial workers, but each of them is more important, not less important, than eight years ago.

It does not seem like this in many parts of the country because factories which were familiar landmarks for generations have closed in the last decade, but the firms which own them have not disappeared or decamped abroad in most cases; they have simply concentrated production in fewer factories, often investing more than they used to in them. The picture is somewhat similar in very many of the individual factories that have survived. Workforces have shrunk. But output has increased, and each factory can be more important to the firm as a whole than a decade ago. Look, for example, at what has happened at Ford Dagenham. The company’s director of industrial relations tells:

“We have been investing heavily in new technology. If you go into Dagenham nowadays it is quite different from 10 years ago. If you go into the body plant you wonder where all the people have gone. Once it was full of men hammering away, with spot-welding guns hanging from the roof. We have invested £1,600 million since 1979 in Britain at the same time as reducing the labour force by 40 per cent, and we have another £430 million in the pipeline.” [12]

He boasts that Ford Britain ‘is the only mass producer of cars in Britain that has consistently made a profit’. He could have added that this makes it one of the few mass producers of cars anywhere in Europe to have done so!

A similar picture is to be found even in Britain’s oldest industry, textiles. Here there has been a huge drop in employment of about a million workers since the 1950s. And the decline is continuing. Yet it does not always mean a fall in output. As the Financial Times noted:

‘A recent investment decision by Courtaulds, the textile group, provides an insight into the trend ... The company is spending £4.5 million to re-equip its Maple No.2 spinning mill in Oldham near Manchester with the latest technology – in this instance, Schlaforst Autocoro machines ... Labour productivity at the mill is expected to be more than double. And since Courtaulds does not foresee an increase in the factory’s output, employment is to fall more than half from 259 to 100 workers’. [13]

A questionnaire to stewards in 1984 revealed that output had risen in the previous five years in the majority of workplaces in the print, chemicals and food and drink; even in engineering, where it had fallen in 49 per cent of workplaces, it had risen in 42 per cent. [14] In manufacturing industry as a whole there are fewer workplaces than a decade ago. The decline has been distributed right across the range of workplace sizes, although the bigger ones have fallen in number proportionately more than smaller ones. But this is a long way from there being a large-scale tendency for small factories to proliferate at the expense of large ones. Rather, the tendency right across the board has been for firms to concentrate production, in about half the larger factories and about two-thirds of the intermediate-sized ones.

It can be seen from the figures that just under a quarter of all manufacturing employees are still in workplaces of more than 1,000, and that 40 per cent are still in workplaces of more than 500. Only about 25 per cent are in workplaces of less than 100. These figures refer to total employees, white-collar as well as manual, managers and technicians as well as workers. And they exclude part-time, temporary, contract and some warehousing labour. But they give some rough indication as to what has been happening to manual worker concentrations. And they suggest those concentrations have not been changing all that much.

Manufacturing establishments employing more than 19 workers by size [15]













































(Estimate of number of workers in factories of less than 20: 620,000)

The point is important when it comes to levels of working-class organisation and consciousness. For there is a lot of evidence that it is in bigger factories rather than very small ones that workers have the strongest union organisation. So, for instance, a study made in 1980 (after a year of a relatively high level of industrial action) showed the following pattern for strike activity in different sized workplaces:

Number of manual workers in establishment [16]









% of establishments
reporting industrial action








But it does not follow from these figures that a reduction in workplace size automatically produces less militancy. For not all industrial action is by the whole workforce: you would expect more sectional actions in bigger than small workplaces if only because there are more sections. In particular, there does not seem much change in die frequency of action as workplaces increase from 500 to 1,000. In addition, as the authors of the study put it, ‘it is surprising that establishments with as few as ten or twenty manual employees should be affected by industrial action to the extent shown by these results’.

Finally, the survey also found that in private industry the proportion of workers in the union did not depend only on the size of the workplace, but also on the size of the enterprise. And if one feature of the economy since 1973 has been a fall in the average size of workplaces, another has been a growth in the number and size of large enterprises through takeovers and mergers.

These provisos mean that a fall in workplace size has not necessarily produced an irreversible decline in union organisation and workers’ militancy. In fact, there are good reasons for thinking otherwise. It may be easier in the first place to organise a strong union in a big rather than a small workplace. But once in existence, the traditions of organisation and militancy do not simply disappear with a reduction in its size. This is confirmed by the survey of Batstone and Gourlay in 1984. Their findings suggested that ‘there has been no fall in union density in organised plants’. [17] The number of stewards has fallen less than the number of workers, reducing slightly the number of workers each steward has to represent, and most stewards’ committees continue to meet at least monthly. [18]

This picture is confirmed by a report for the Engineering Employers Federation, based on a survey of 60 workplaces in 1984, which showed that 82 per cent of manual workers were in unions. The study even concluded that there had been ‘a consolidation of trade union membership’ since 1969. [19] Both surveys suggested that a third of private manual workplaces experienced industrial action in 1983-4 and that the big majority of managements did not have any perspective of trying to destroy workplace trade unionism.

The manual service working class

People often equate the manual working class with the production industries, and white-collar workers with ‘services’. The conclusion is then drawn that the growth of service employment compared with manufacturing employment involves a ‘decline’ in the working class. But the equation is quite misplaced. Some of the most important ‘service industries’ employ overwhelmingly manual workers of the ‘traditional’ sort. Dustmen, hospital ancillary workers, dockers, lorry drivers, bus and train drivers, postal workers are all part of the’ service’ workforce. And a very big part.

Total employment in ‘services’ was 13,436,000 in March 1985. ‘Distribution, hotels, catering and repairs’ accounted for 4,240,000 jobs, ‘transport and communication’ for 1,263,000 jobs, postal services and telecommunications for 400,000 jobs, refuse disposal and ‘cleaning services’ for 293,000 jobs, laundries, dry cleaners, hairdressers etc. for 175,000 jobs, and hospitals, nursing homes and so on for 1,307,000 jobs. Nearly 60 per cent of all ‘service’ employment is covered by these categories. And each of them includes quite large numbers of stereotyped ‘traditional manual’ jobs.

So manual workers are still about half the employed workforce, despite the decline in manufacturing employment.

I have not been able to find any detailed breakdown of this section of the manual workforce according to workplace size. But certain things we do know.

  1. Some traditional groups of ‘service’ manual workers have been in decline for many years. So the number of registered dockers has sunk massively in the last three decades from 70,000 in 1956 to 40,000 in 1970 [20] to about 14,000 today. There has been a somewhat smaller decline in the number of railworkers, from around 425,000 in 1951 to about 147,000 in 1985. This decline has meant both a fall in the absolute number employed and a fall in average workplace size.
  2. Some groups have shrunk with ‘rationalisation’, but much more slowly. This applies, for instance, to road transport workers. After an increase in number from 300,000 in 1925 to over half a million in 1951, their numbers fell to 461,000 in 1974 and 383,000 in 1985. But in these cases there has not necessarily been a fall in the average size of workplaces (such as transport depots). It is quite likely that the process of rationalisation which has reduced the workforce has also led to a growing domination by larger firms and, possibly, larger workplaces.
  3. Some groups have hardly changed in size at all recently. So, for instance, there were 431,000 workers in ‘postal services and telecommunications’ in 1976; there were 420,000 in 1985. The number of workers in refuse collection has only fallen from around 80,000 to about 70,000, while the number in ‘cleaning services’ has remained fixed around the 210,000 mark.
  4. Some groups of manual workers have grown substantially, at least until recently. For instance, the number of workers in the health service doubled between 1951 and 1974, and has continued to increase, although slowly, since. This increase has included an increase in the numbers of manual workers. There has been a similar upward growth in areas such as hotels, and catering (where the workforce has risen by 11 per cent in the five years up to 1986 to just over 1 million) and in retail distribution (where it has grown by about 6 percent to 2,270,000).

One result of these changes is that although the union membership among some groups of manual service workers has fallen with the fall in employment (for example dockers and railworkers), in other areas union membership grew massively until recently:

Union membership (000s)







National Union of Public Employees






Union of Postal Workers/Communications Workers






In 1980, two-thirds of the General and Municipal Workers Union’s 900,000 members were in public sector ‘services’ and one-third of the Transport and General Workers Union’s 1.9 million members.

The overall shape of the ‘traditional’ working class

Most discussion on class in Britain breaks down the population by occupation according to the criteria employed by the government’s statisticians. [21] On this reckoning, manual workers made up 51.8 per cent of the 20,890,000 employees in 1980. [22] The destruction of some manual jobs since then means that today manual workers are less than half of all employees. Nevertheless, they still account for more than 55 per cent of male workers and about 35 per cent of female workers. [23] These figures are an important corrective to those who see the traditional working class as disappearing. But the classification by occupation obscures some important distinctions among manual workers: in particular, it includes foremen along with ordinary workers. Two surveys of class have attempted to remove this distortion. The first is by Heath, Jowell and Curtice. [24]



% of total population



Self-employed and small employers



Foremen and technicians



Working class



The second, by Goldthorpe and Payne [25], is of males in the age bracket 31-75. It concluded that ‘lower technicians and manual supervisory’ were 11.5 per cent of the group, ‘skilled manual’ were 25.3 per cent, and ‘semi and unskilled manual’ were 26.5 per cent

The lack of agreement between these two surveys shows the real difficulties of coming to a clear picture. The difficulties arise because there is no clear-cut boundary between the ‘manual working class’ and other groupings, such as routine white-collar workers (so Heath, Jowell and Curtice make this group twice as big for male workers as do Goldthorpe and Payne) and the self-employed (many ‘self-employed’ manual workers are, in fact, on the ‘lump’ and should really be counted as employed workers).

Yet the authors of the two surveys come to some common conclusions that are quite important. The first is that there are certain objective features of manual working-class life that remain unchanged – indeed, which might even have been accentuated – despite the shrinkage in the size of this grouping. Thus Heath, Jowell and Curtice argue that ‘manual wage labourers have relatively little security of employment and relatively poor fringe benefits such as sick pay and pension schemes. They have little control over their own working conditions and little discretion over what they do at work’. [26]

Goldthorpe and Payne point out that unemployment affects about one in five of the manual working class. Department of Employment figures show that semi-skilled and unskilled male manual workers are more than twice as likely to be affected by unemployment as the population as a whole, while the ‘professional managerial’ group only suffer about 40 per cent of average unemployment. [27]

Other figures bear this out. The British Travel Survey, for instance, shows that 56 per cent of semi-skilled and unskilled manual workers did not have a holiday at all in 1984, while for the professional-managerial group, 20 per cent had three holidays, 20 per cent two holidays and 40 per cent one holiday. [28] While 44 per cent of council tenants in the ‘professional-managerial group’ had considered buying their own house in the last two years, only 27 per cent of semi- and unskilled tenants had’.

There is little mobility in the manual working class from other groupings – 70 per cent of male manual workers had fathers who were manual workers and only 7 per cent fathers who were in non-routine white-collar positions. [29] And working-class children have few chances of climbing into the ‘professional-managerial’ group via education: in 1984 70 per cent of new university students came from the ‘professional-managerial group’, 12.4 per cent from skilled manual’ (including foremen and self-employed manual), 6.2 per cent from ‘partly skilled’ manual and 1.1 per cent from unskilled manual. [30] So of more than 450,000 children from manual working-class backgrounds, only about 12,500 were accepted at university, add the numbers for polytechnic and other sorts of higher education, and you still find that only about 5 per cent of working-class children can ‘rise’ out of their class in this way.

Third, although the traditional manual working class may be shrinking, it is still far from being absorbed by the cultural traditions and ideology of the middle class. Heath, Jowell and Curtice conclude that their figures ‘do not confirm a decline in relative class voting in 1979 and 1983. The level of class voting is not evidence that subjective class voting has declined any more than objective inequalities have’. [31]

The cultural separation of the manual working class is shown by newspaper readership. Although all classes in Britain read capitalist newspapers, the manual workers read different papers from those of the upper and middle classes. ‘While about a third of manual workers in Britain read the Sun and about 25 per cent the Mirror, the proportion of manual workers reading all the ‘posh’ papers (the Telegraph, the Guardian, the Times and the Financial Times) combined is well under 10 per cent – and about 30 per cent read no daily paper at all. [32]

Even the most notoriously unreliable of all figures on social class – those on which class people ascribe themselves to – seem to confirm this picture. In 1984 48 per cent of people rated themselves as ‘working class’ and 19 per cent as ‘upper working class’. [33]

The growth of a ‘peripheral’ ‘servant class’

The figures so far should help to dispel the Gorz myth of the ‘disappearance’ of the manual working class. But some of them would be quite happily accepted by the revamped ideas of those who follow Hobsbawm, which base themselves on John Edmonds’ talk of a ‘new servant class’ and on theories of ‘core’ and ‘periphery’ workforces.

These theories centre round the notion that although the old manual working class in production and in certain old service industries remains intact, it is diminishing in size for two related reasons. First, they point to the decline in the number in production industries and in many of the old service sectors where trade unions are well-organised, and to the rise of sectors such as hotels, catering and distribution which are very badly organised. As the labour editor of the Financial Times rather crudely puts it:

‘Britain’s unions have failed to crack these new industries – failed to follow the work. Employment has shifted massively from manufacturing, hit hard by the recession, to the private sector service industries ... such as tourism, hotels, and restaurants ... The unions have not followed suit.’ [34]

Edmonds expresses essentially the same idea:

We must accept that within the next decade the trade unions are not going to be in a position to force contract cleaners, for example, to pay reasonable pay and conditions through traditional trade union organisation. We are not going to have effective trade union organisation in every large hotel in the country ... The whole private service area, particularly leisure, isn’t well organised and is likely to remain significantly unorganised for all sorts of structural reasons ... If you have an industry where the workforce is highly mobile, where they are not attached to any particular employer for any length of time, then the organisational difficulties are very substantial indeed. It is obviously more difficult to organise there than it is in a factory of 500 people who have relatively long service records. [35]

There are a number of obvious points that can be made about Edmonds’ claims. Groups such as contract cleaners could often easily be organised if the unions in organised workplaces would black non-union firms; they do not do so because people like Edmonds are too cowardly to confront the Tory laws.

Over the years union activists have succeeded in organising industries such as the print where workplaces are, on average, much smaller than in the big hotels or the big chain stores: the big London hotels employ two or three hundred workers apiece, and 60 per cent of Tesco’s stores employ more than 100 people, and 30 per cent more than 200 people; while in the very well-organised printing industry 96 per cent of firms employ fewer than 100 people and account for half the industry’s workforce. [36]

It is not true that all hotel, catering, distribution or ‘leisure’ employees are continually changing jobs; usually there are groups A permanent workers (chefs, porters, cashiers, warehouse personnel) who can maintain continuing organisation, as is shown by the fact that there have been successful attempts at unionisation.

But there is a more central fallacy than any of these in Edmonds’ approach: the fact that he concludes from the failure of the unions to build in such a growth area that therefore the union strength is bound to decline generally. Union strength does not just depend on how many workers are organised, but on the ability to take economically effective action. And even if the hotel, catering, retail and leisure sectors remain weakly organised, other sectors do retain mass potential strength.

The concentration on numbers is a prime concern of union officials worried about how to sustain the subscription income which keeps the union bureaucracies ticking over. It cannot be of the same sort of concern to socialists aiming to analyse scientifically the strength of working-class organisation.

Core and periphery? Much more important in some ways is the argument about the core and the periphery. For this suggests that the unorganised ‘new servant class’ is not confined to an economically less important sector of its own, but is being used by employers to undermine the central bastions of trade union strength. The claim is that firms have responded to the recession by reducing the number of permanent, full-time, unionised workers and increasing the number of part-time, temporary, and contract workers.

As Atkinson and Gregory put it: ‘The most important changes have involved the reorganisation of firms’ internal labour markets and their division into separate components, in which the worker’s experience and the employer’s expectation of him/her are increasingly differentiated.’ [37] ‘Employers hope to achieve flexibility’ with the core workers ‘by stressing mutual long-term commitment’ while with the periphery, ‘exposing workers more and more to raw market forces’. It is clear that the development of an employers’ offensive ... is likely to restructure the labour market in ways which have the most profound consequences for workers and their trade unions ... At their simplest [the changes] boil down to security and incorporation for the few, bought at the expense of deterioration in conditions of employment for the many. [38]

If this were indeed happening, it would clearly have very dangerous implications indeed for the strength of the working class. But the evidence suggests that it is not happening, and cannot, short of a prior catastrophic defeat for the whole trade union movement.

The proponents of this theory point to a number of things which have happened in recent years – the growth of part-time employment, particularly among women, an increase in the number of casual workers, and tendencies in certain industries both to contract out certain work, and to employ contractors to do certain internal tasks. All these trends do exist. And in a few cases they have been used to restructure workplace employment in the way the theorists suggest. But these cases are few and far between, and are likely to remain so.

The growth of temporary labour is a response of companies to a situation in which a recession has cut the size of the workforce and in which there is a growth in demand which may not last. The growth of part-time labour is a response to a situation when the expansion of output is not likely to be big enough to justify taking on full-time workers. But neither change is likely to last indefinitely. If output rises permanently, then the firm will require a stable, reliable permanent workforce and is likely either to expand the workforce – turning temporary workers into permanent workers and part-time workers into full-time workers – or to increase output from existing workers by introducing new work systems (for instance, round-the-clock shift working) that put more pressure on the existing ‘core’ workforce.

As Batstone and Gourlay point out: ‘The logic of resort to secondary labour is to have flexibility. If this is so, there are pressures on employers to meet falls in demand by shifting work away from the secondary sector into the primary, unionised sector. Therefore the question arises as to whether at a particular time in a recession, employers are shedding labour rather than recruiting it’. [39] Incomes Data Services Focus suggests an example:

Manufacturers in the food industry, undertaking heavy capital investments, may no longer operate with the inefficiencies of the ‘twilight’ or split shift but move to 24-hour operation with seasonal peaks. Insofar as they employ ‘peripheral’ workers at all, the change will come within this group. Part timers will become seasonal workers or disappear. [40]

The closer you look at the argument about core and periphery, the more holes you find in it. For instance, there are firms in which part-time workers have permanent contracts: the employer wants to be guaranteed the attendance at work of the part-time worker for the foreseeable future because he wants to maintain a regularity of production. Such part-time workers can be unionised, as they are in the food industry. [41] They ‘can be treated as primary rather than secondary labour’. [42] And there are even firms in which the employer wants ‘regular casuals’, workers he knows will always be available to satisfy seasonal increases in demand. Again, contract labour is by no means always non-union labour. Some important contracting firms are themselves very big, with their own ‘core’ of unionised labour.

What is more, there are limits within which firms are going to rely on contract labour, unless they themselves dominate the contracting firm. As the Incomes Data Services article puts it:

In reality many employers are finding that subcontracting firms need to be kept under close control and in a permanent relationship. The employees of these firms are very much of the ‘core’ because the functions they perform – notably maintenance – are critical. Other firms would not dream of letting maintenance out of their control for a minute: they contract out the traditional functions like catering, haulage or security. [43]

In areas such as computer maintenance it can be shortage of skilled labour, rather than any strategy for breaking unions, that leads firms to accept the risks involved in resorting to outside contractors who cannot always be relied on to be available when required.

The survey by Batstone and Gourlay suggests that in fact part-time and casual labour has been much more a feature of non-manual rather than manual industry. ‘Subcontracting has become more important on the engineering side’. [44] ‘Employers in manufacturing have typically increased labour input through greater and more intensive use of their primary labour force than through widespread use of secondary labour’. And even where secondary labour has been used it has also been accompanied by ‘an extension of the hours of primary labour [through shift and overtime working]’. [45] For instance, increased use of temporary and casual labour in the civil service clerical grades has gone alongside a massive increase in the workload of the existing workforce.

This points to the most important error with the core-periphery theory – the belief that employers can somehow use secondary labour to give a privileged position to ‘primary’ workers. In a context of national and international economic turmoil, there is no way many employers can do this.

Interestingly, at one of the few places where a real division between a ‘core’ and a ‘periphery’ workforce has been imposed, at the UIE shipyard on Clydeside (formerly John Brown), a recent dispute was over management’s announcement that it was sacking half the ‘core’ of ‘permanent’ workers. As the IDS article rightly remarks: ‘This year’s core group become the casualties of next year’s downturn. There is no way most firms can draw a chalk circle round a section of the workforce and guarantee jobs for life’.

The use of casual instead of permanent workers is one ploy that firms will try in an effort to weaken union organisation. As such trade unionists have to resist it. But it is not an all-embracing strategy that the employing class is successfully pushing through in a way that can be expected to produce a new, permanent divide in the working class.

Proponents of the theory that it is are making a mistake which mirrors the claim by right-wing trade union leaders such as Eric Hammond that a Japanese ‘dual labour market’ can be successfully introduced in Britain. They are claiming that a pattern of employment that developed in Japan in the aftermath of the Second World War can be imposed successfully in Britain today. But conditions are very different. The Japanese employers faced a workforce whose attempts to organise had just been smashed in the face of 66 per cent unemployment, and could benefit from the market provided by a world boom that was to last nearly 30 years. In Britain today, employers face trade union organisation that still has very strong roots, they continually complain that unemployment has not destroyed all shop floor resistance, and they cannot hope for any sustained world boom.

The reality of manual working-class organisation in Britain today is much better expressed in a recent report by the Policy Studies Institute:

‘The unions retain a solid base of membership and organisation over wide sectors of British industry, based characteristically on self-reliant workplace union organisation with a high membership density, often supported by closed-shop provision, strong convenor and senior steward leadership, based upon a close familiarity of the site’s activity and a good understanding of what members will expect and will offer and a set of – often informal – procedural rights based on custom and practice’. [46]

The manual working class in Britain has long been characterised by a marked cultural separation from the rest of society and strong defensive trade union organisations on one hand, and acceptance of key elements of ruling-class ideology on the other. The expression of this combination has been the hegemony of right-wing Labourism, challenged on the left by a small minority of socialist activists, and on the right by a substantially larger, although less active, minority of workers committed to ‘Alf Garnett’ Tory-type views. [47] By and large that pattern continues to prevail today.

The white-collar working class

One of the major trends cited by those who write off the traditional’ working class has been an almost continuous increase in the numbers of white-collar employees for almost half a century.

For the cruder analyses the white-collar workforce is seen, in its entirety, as part of the middle class. But in fact the growth of white-collar work in this century has been accompanied by a reduction in the differential of average white-collar salaries compared to many wages, until by the 1950s the ‘level of remuneration of white-collar workers was ‘comparable to that of skilled labourers’. [48] But the change in average differentials conceals something else which occurred simultaneously. There was a growing differentiation within white-collar work, with a layer of highly-paid executive positions emerging on the one hand, and a mass of lower-paid routine non-manual jobs on the other.

As the French sociologist Crozier has noted:

‘A split has taken place between highly qualified employees charged with handling matters demanding judgement, experience and responsibility, and a mass of unskilled employees assigned to a series of simple, unchanging operations. In the administrative services of banks, insurance companies or large accounting firms there have for some time been numerous cases of assembly line work ...’ [49]

This polarisation within white-collar employment took place at the same time as the entry of large numbers of women into white-collar work. The proportion of office employees who were female went up from 0.8 per cent in 1851 to 59.6 per cent in 1951. [50] Also, ‘The arrival of women ... was accompanied by a process of mechanisation and automation whose effects on males were thereby diminished. The latter were pushed towards the more skilled occupations and towards executive positions, so the general proletarianisation of the white-collar group was not experienced as such by those concerned.’ [51]

So in France, ‘To the 600,000 male employees of 1920 there now correspond probably 350,000 supervisors and 250,000 highly qualified employees whose status is at least equal to that of their predecessors in 1920’. [52] Overall: ‘To the old white collar group, which had pretty much retained its social status, was added a new group consisting in part of females with a distinctly inferior social status ...’ [53]

For this reason, most modern sociological accounts of white-collar employees divide them into two main groups: the professional-managerial (sometimes called the ‘service class’) [54], and the clerical and other routine non-manual. Most estimates of the professional-managerial ‘service class’ claim it makes up ‘some 20-25 per cent of the population’ [55] and is growing quickly. So for Britain the Labour Force Survey of 1981 put 24 per cent of all employed people in the managerial and professional category, 17.4 per cent in the clerical and related category, and 7.7 per cent in the ‘other non-manual’ category. [56] The ‘clerical and related’ workers are clearly part of the working class in general. The category includes, for example, ‘retail shop cashiers, check-out and cash and wrap operatives’, ‘typists, shorthand writers, secretaries’, ‘office machine operators’, and even ‘petrol pump and forecourt attendants’ and ‘firemen’. [57]

The pay and working conditions of clerical workers have grown closer and closer to those of the mass of manual workers over the last half century:

‘By the First World War male clerks and skilled manual workers earned the same amount, and this parity is maintained until about 1936. Between 1935-6 and 1955-6 the trend changes quite markedly against clerical workers ... By 1970 there had been a further narrowing of the gap ... By 1978 the average earnings of semi-skilled men had overtaken their clerical counterparts for the first time ... In addition, many of the traditional superior employment conditions of clerks have been gained by manual workers’. [58]

The proletarian character of clerical and similar white-collar workers is further proven by the fact that many of them are the wives, sisters and daughters of male manual workers. A third of clerical workers come from manual working-class backgrounds, a third from clerical backgrounds, and only a third from the so-called ‘professional-managerial service class’. [59]

However, there are various provisos to be made before simply equating routine white-collar workers with manual workers.

Firstly, the massive rise in white-collar trade union membership over the past 40 years cannot simply be equated with a growth in unionisation of the lower more ‘proletarian’ grades.

Of the growth, however, there is no doubt. The proportion of white-collar employees in unions rose from 21.1 percent in 1911 to 32.6 per cent in 1968 to 39.5 per cent in 1974, to 43.0 per cent in 1977 to 44.0 per cent in 1979. [60] While the manual unions have often suffered a loss of membership with the recession, some white-collar unions in the public sector have continued to grow. In engineering, national government and local government the fall in union membership was 3.8 per cent, 2 per cent and 1.7 per cent respectively in 1983-4, but in health and in education it went up 2.4 per cent and 3.4 per cent in the same period.

But this has not necessarily been due to an increase headed by female clerical workers. A study of union membership in a bank, an insurance company and a local council concludes that women’s union membership was about half men’s: ‘In all three institutions the level of union organisation rose with grade level’. [61] Nor is there ‘any significant relationship between women’s social origins and union membership’. [62]

When it comes to union activists, a study of the white-collar employees of Sheffield council in the late 1970s showed that although 54 per cent of the union members were in ‘low status’ jobs, 70 per cent of the shop stewards were in ‘medium’ or ‘high status’ jobs. [63]

Secondly, the growth in white-collar unionisation still has not created the same traditions that exist among manual workers. The level of unionisation generally is lower than that which exists in manual industries. The union density of white-collar workers was 35 per cent in 1974; this compares with 52.1 per cent for all manual workers. [64] In engineering in 1984 the comparable figures were 48.5 per cent and 82 per cent respectively. [65]

In terms of political affiliations there is a significant difference between routine white-collar workers and manual workers. On average only 35 per cent of routine white-collar workers have voted Labour in the last seven general elections (and only 25 per cent in 1983), as against 42 per cent who have voted Tory (46 per cent in 1983). [66]

How are these differences to be explained?

First, some routine white-collar workers do have a very real chance of moving out of their lowly jobs, in a way in which manual workers do not. White-collar work is typically organised hierarchically, with low-level employees being able to seek promotion to higher grades if they can pass various internal or external exams and please those above them. The pyramid shape of the hierarchy, with far fewer jobs at the top than the bottom, means, of course. that only a small number of employees can rise up like this.

The women who make up the big majority of low-level white-collar workers have few chances of promotion – because of managerial prejudices against them and, more importantly, because the impact of marriage and pregnancy has been to take them out of their jobs before the point at which promotion would be open to them. So while it is not true simply to say, as some feminists do. that ‘women are the low paid’, only 10 per cent of non-manual women earned the £173 or more a week that half non-manual men got in 1984. [67]

This has necessarily blunted the pressures towards unionisation and militancy. The male workforce has been made up men who can expect a fair chance of improving their conditions by looking towards promotion rather than collective action.

Not all male clerical workers, however, attain upward mobility. In one bank head office, for example, although 88 per cent of the men over the age of 35 had achieved some degree of promotion, this was only after half the men who started on clerical grades had dropped out by the time they reached their late 20s. [68] One study suggests that 26-31 per cent of men who started off in their working lives in clerical work end up in manual employment. So the proportion getting eventual promotion is only about half. And it can take many years before promotion comes. ‘Many–probably a majority – of men spend a considerable proportion of their working lives waiting before they achieve unambiguously managerial positions, if they achieve them at all.’ [69]

But the eventual success of half the male clerical workers can have profound effects on the attitudes of the whole workforce. As Crozier has noted:

‘The [male] white collar worker does actually have a better chance than manual workers] to attain, or see his children attain, higher posts. If he models himself on the upper class it is not simply because he finds himself subject to the influence of the rulers ... it is also above all because he is well aware that they have a tendency to choose for eventual promotion those persons whose behaviour comes closest to their own standards. To be sure, the white collar employee is at the bottom of the ladder, but at least he is on the ladder.’ [70]

This attitude of the male employees inevitably exercises an influence on female employees as well. Surveys of French office workers showed how office workers become involved ‘in the organisation system of which they are part’:

Subordinates do not seem to form a world apart, but are deeply affected by the hierarchical system to which they belong. The behaviour of upper management is the determinant factor for executives’ leadership style, and this in turn is a decisive factor for personal relations in general and for atmosphere ... Contrary to the blue collar group, separated and protected as much as oppressed by die class barrier, white collar workers do not respond by means of solidarity, but rather by indifference and apathy ... Indeed, Crozier argues, even when low-level white-collar workers are actively hostile to their superiors, this can well be because the superior is not active or efficient enough in organising the work that has to be done – that he is not a ‘forceful and authoritarian supervisor’. [71]

Again, white-collar workers often work in relatively small groups, even when employed by very large concerns. These groups will include employees at different levels up the hierarchy. For example, in Sheffield council, 39 per cent of stewards represented employees in widely-scattered work sites, and only 31 per cent represented a workforce of ‘homogeneous’ status. ‘61 per cent have members covering the full range of statuses within departments’. [72] The presence of ‘bosses’ in small work groups is clearly likely to blunt expressions of collective, class feeling much of the time.

It can be suggested that there is another factor at work influencing the ‘atmosphere’ in routine white-collar work. Although few women employees have an opportunity for upward mobility through careers (although there are some indications that this is changing) [73], a minority of them will be able to enjoy the living standards of the ‘professional-managerial’ or petty bourgeois groups through marriage to men higher up the hierarchy (or to men who intend to make a career themselves). This in turn may lead others to identify with the groups above them in the hierarchy rather than to think in collective, class terms. [74]

Finally, the oppression of women outside the workplace affects the organisation of routine white-collar workers in much the same way that it affects the organisation of female manual workers. The unmarried women have not, traditionally, expected to be working in the job for more than a few years, and therefore do not under normal conditions show a great interest in union organisation. And the older women, who return to work as their children reach school age, are often impeded from playing a central, organising role in the workplace because of the continued responsibility for childcare (which forces them into part-time jobs and which limits their ability to go to after-hours union meetings). So, for example, despite the fact that 53 per cent of Sheffield council’s white-collar workforce were women, only 26 per cent of stewards were women. [75]

In manual industry women workers have generally in the past come to the fore when the class as a whole is moving forward into struggle. Then their double oppression can act as a spur driving them to great levels of militancy. But when the class as a whole is in retreat, it can have the reverse effect, of being a fetter which prevents them playing a full and effective part in sustaining low-level organisation.

None of this means that routine white-collar workers cannot be organised. If nothing else, the experience of the clerical union in the civil service, the CPSA, proves otherwise. A survey in 1984 concluded that ‘some form of industrial action’ had taken place in ‘all but a handful’ of civil service offices, and tended ‘to be concentrated in the smaller establishments’. [76] Yet, even in that union, the attraction to some of the ablest militants of upward mobility and the continual turnover in staff presents a big obstacle to sustained rank-and-file-led organisation.

Most groups of routine white-collar workers can be expected to follow, rather than lead in upsurges of working-class struggle. They are a key sector of the working class, but are usually likely to be dependent upon the more ‘traditional’ groups of manual workers to take the lead in class-wide struggles. And when it comes to routine union organisation, the initiative is often not taken by them, but by those a step above them in the bureaucratic hierarchy.

The new middle class

A growing proportion of the working population is made up of white-collar employees above the clerical rank. Of this there is little doubt. The Department of Employment puts 34.5 per cent of employed males and 25.5 per cent of employed females into the ‘managerial and professional’ category. But what is the social character of these employees? Are they all members of some new class – a ‘service’ class or ‘salariat’ which might even be as big as the manual working class – as some fashionable sociologists claim? [77]

Marxists do not generally accept terms such as ‘salariat’ or ‘service class’. These are based on a notion of class derived from Max Weber in which what matters is opportunities for living a certain lifestyle, rather than on relations to the means of production. However, Marxists have had to try and develop their own analyses of the higher white-collar grades. [78]

Again and again socialist activists have found that when the members of these grades attain trade union positions they use them for managerial purposes as much as for the defence of the conditions of other workers. This is confirmed by a survey of Sheffield local government workers. It was found that ‘48 per cent of stewards described their work roles as managerial’ and that 81 per cent of these ‘claim they experience “two hats problems” of inter-role conflict’ between being stewards and being managers. [79] In the case of the civil service, socialist activists in the clerical union, the CPSA, have opposed a merger with the union for executive grades, the SCPS, on the grounds that it would mean having managers at union meetings.

The best way for Marxists to come to terms with these problems is to recognise that a minority of those above basic clerical grades make up a ‘new middle class’. The term was first used by Kautsky and then by Trotsky, and its use has been most recently developed by Alex Callinicos.

This is the class of those who hold the higher and better remunerated positions in burgeoning bureaucratic structures that characterise ageing capitalism. They are distinct from the top bureaucrats who actually run these structures (who are virtually indistinguishable from the ‘private’ entrepreneurs who run capitalist firms; they are equally committed to the exploitation of the working class and are therefore ‘personifications of capital’, in Marx’s words). But the new middle class is also distinct from waged workers, whether manual or white-collar. Waged workers get no more than the value of their labour power, whereas the members of the new middle class get an income considerably greater than the value of their labour power or even than the value their labour would create if applied in productive industry; in this way they can even gain from the exploitation of waged labour. [80]

This does not mean the new middle class automatically identifies with capital. Like the old middle class or petty bourgeoisie of small employers and self-employed professions, the new middle class finds itself in a contradictory situation. It occupies a subordinate, dependent role in the system. Particularly in times of social crisis it may risk being ground right down by the ruling class (the petty bourgeoisie risk bankruptcy, the new middle class redundancy). Yet at the same time it gains considerable privileges from aiding and abetting the ruling class in controlling and exploiting the working class (the petty bourgeoisie does so directly by exploiting its own employees, the new middle class in so far as it ‘earns’ high salaries by helping to impose the needs of accumulation on those below it).

The boundaries of the new middle class cannot be delimited in any absolutely clear way, any more than can the boundaries of the old petty bourgeoisie: at the top of the bureaucratic hierarchy it blurs into the class of managerial capitalists, at the bottom into the white-collar working class. To this extent it is not an independent class, but one that is pulled this way and that, depending on the pressures on it. But this makes some rough delimitation of its boundaries all the more important, as the only way to understand the links which bind certain elements in white-collar trade unionism to management.

The easiest way for Marxists to proceed would be simply to adopt the Labour Force Survey’s ‘managerial and professional’ categories, Heath, Jowell and Curtice’s ‘salariat’ or Goldthorpe’s ‘service class’ and call them the ‘new middle class’. But it would be a profound mistake to do so. These categories all contain a majority of people who do not exercise any real managerial power over those below them or receive salaries greater than the value of their labour power. Thus the managerial-professional category in the Labour Force Survey includes such groups as all librarians, all teachers, all nurses, radiographers, physiotherapists, medical technicians, journalists, window dressers, actors, singers, lab technicians and supervisors in shops and offices. Most of the people in all these occupations are at or relatively near the bottom of bureaucratic hierarchies.

It hardly makes sense, for instance, to put all female nurses in the ‘managerial and professional’ category, when their average pay in 1984 was slightly less than that of ‘secretaries and shorthand typists’. Again the average wage for all secondary teachers (including in this the most highly-paid heads) was only about 25 per cent more than the average male manual wage. [81]

A number of indicators suggest that the new middle class only includes about a third of those in supposedly ‘managerial-professional’ grades of white-collar work.

(i) Income. Only about a quarter of non-manual men – about 10 per cent of the male workforce – earned more than twice the average manual wage of £153 in April 1984. The average rate of exploitation in Britain in 1986 is well over 100 per cent. So someone earning twice the average wage may be less exploited than the average, but is certainly not living off the exploitation of others.

With women it is even clearer. Half of women white-collar workers earn less than £120 a week, and only 10 per cent more than £190. So it would be nonsense to put all the 25 per cent of women in allegedly ‘managerial and professional’ jobs into the new middle class.

(ii) Qualifications. Increasingly, the precondition for higher positions in non-manual employment is some sort of formal qualification. Yet of people in the age group 25-40, only 10 per cent have degrees or the equivalent, and only another 10 per cent other higher education qualifications (HNCs, teaching certificates and such like). Now not all these get higher white-collar jobs, and not all those who go into these jobs have such qualifications. But the figures again suggest that not more than half the ‘managerial and professional category’ can be so different to other employees as to merit inclusion in the new middle class.

(iii) Studies of particular work situations. Crompton and Jones show in their study of three white-collar employers that promotion out of clerical grades does not usually mean promotion into truly managerial positions. ‘The administrative and managerial’ grades are now set at ‘lower levels of the organisational hierarchy than hitherto’. [82] The result is that the majority of men who are promoted never take on managerial functions, but simply do ‘superior clerical jobs’. [83] And while 12 per cent of women were in ‘supervisory grades’, only 1 per cent achieved managerial positions. [84]

Nicholson, Ursell and Blyton seem to apply similar criteria for the union membership among white-collar employees of Sheffield council. They break them down into three groupings, low, medium and high, rather than just into routine and non-routine. They put 54 per cent of the workforce in the low category, 30 per cent in the medium category, and only 16 per cent in the ‘high’ category, with its clear managerial functions.

There is one analysis of the ‘service class’ – the most recent one by Goldthorpe and Payne [85] – which does break it down as our figures and these studies suggest it should be. At one point they distinguish between ‘upper’ and ‘lower’ ‘professional/administrative/ managerial’ categories (only to then abandon this distinction and to merge the two groups). The upper grouping contains only 9.7 per cent of men between the ages of 31 and 75. This would seem to coincide with the sort of rough estimate for the size of the ‘new middle class’ suggested by the figures for incomes and qualifications, and by the case studies on managerial authority. It is significant as a proportion both of the total population and of the white-collar workforce. But it is a far cry from anything approaching the traditional manual working class in size or significance.

Intermediate grade white-collar workers

The account so far shows that the white-collar working class is structured hierarchically to a much greater extent than the manual working class. The supervisory jobs into which manual workers might reasonably aspire to climb – those of foreman and manual supervisors – make up only 2.5 per cent of the total workforce [86] and about one manual job in 12. By contrast, the non-manual ‘new middle class’ accounts for 10-15 per cent of the total workforce and at least one white-collar job in five.

Below the new middle class are a very large number of white-collar employees (about 15 per cent of the total workforce) who, despite being above the basic, ‘routine non-manual’ grades, are workers and are exploited. These middle grades of white-collar work present a perplexing contradiction. They are made up of people who often aspire to promotion into the ranks of the new middle class and have, as we have seen, some chance of achieving it if they conform, work hard and persist in the same careers. But they are also the layer from which come many – in fact, probably most – of the activists who have built white-collar trade unionism.

One study found that ‘levels of union membership rose with grade level’. [87] Another study found that among stewards, ‘female and low status employees are under-represented relative to the membership’. [88] So 41 per cent of stewards came from middle-level jobs, even though these were only 30 per cent of the workforce (and 29 per cent of the stewards came from 19 per cent of the workforce who were ‘high status’). [89]

The authors of this study suggest two reasons for this domination by middle and upper ranks.

First, it is physically much easier to engage in union activities the higher your ranking in the bureaucratic hierarchy. The upper ranks are often required by their job to visit different offices, make phone calls, and so on. By contrast, ‘the lower status steward (typically female) is confined to one location, is often isolated from other workers (63 per cent of low status stewards worked on their own) and has neither the authority nor the means to contact others’. Given that nearly all the stewards found it difficult to find the time to do their jobs and their union work as well [90], this factor must have been of considerable importance.

Second, ‘It was ... frequently apparent that on average senior grade stewards were more self-assured and could deploy greater verbal and communicative skills at union meetings’. [91] In other words, the education and training bourgeois society gives those in the higher grades tends to make them more confident in meetings than those in lower grades. So at stewards’ meetings, ‘the stewards of high job status showed much greater readiness to speak to all items on the agenda’. [92]

A third reason can be deduced from the other study. Lower grade clerical workers have traditionally been women who have assumed that they will only be in employment for a few years before they leave to have children. They therefore have had less interest in achieving long-term improvements in conditions that those who have a foot on the ladder upwards – or at least aspire to have. This would explain one of its most interesting findings: among women employees, trade unionists were 50 per cent more likely to be ‘interested in gaining promotion’ than non-unionists. [93]

But there is an increasing tendency for women employees to see their long-term future as lying in the world of paid labour, and to develop attitudes much closer to those traditionally held by white-collar men. 70 per cent of young unmarried women expressed interest in promotion, as against only 29 per cent of ‘older women in the second phase of their work cycle’. [94]

So the most trade union-conscious white-collar workers are often those with greatest hopes of moving up the career ladder, eventually into jobs in which they will supervise other workers. This explains one of the central peculiarities of white-collar trade unionism: those who are often the most committed union activists, whose activity leads them to play a key role in union branches, are often those who end up in managerial positions.

As already noted, the Sheffield study found that stewards from the upper grades tended to dominate union committees. This affected the way the union functioned: ‘The upper grade stewards were observed disproportionately to exert a moderating influence on ... ends and means, and often drew attention to the constraints under which management were operating ...’ [95] ‘There were signs in the handling of grievances that managerial frames of reference were often imposed’. [96]

This was despite the fact that in terms of their professed politics, ‘the higher status’ stewards were ‘predominantly to the left, while the lower status stewards were predominantly to the right’. [97] The study suggests that many of the higher status stewards were people who had been radicalised politically while students, and then applied their radicalism while climbing the white-collar hierarchy. Clearly, they had ended up combining a quite right-wing practice with some form of vaguely left-wing ideology – which perhaps explains the appeal of Marxism Today-type ideas among such people.

What about the middle grade trade unionists? They certainly do not generally reject aspirations to managerial rank. The trade unionists, it will be recalled, are often those more interested in promotion: one of the reasons the Sheffield study found for activism was ‘desired access to managerial decision making’. [98] And to get promotion there has to be some degree of formal acceptance of the goals of the organisation. Even if activists want to reject such an identification, they can hardly say so in public if they are to hope to rise higher. One reason the middle grades are paid more than low-level workers is to buy such support from them. [99]

Yet, in practice, middle-level white-collar workers rarely exercise any real supervisory, let alone managerial, authority. As Crompton and Jones emphasise:

The work of first-line supervisors differed little from that of their immediate subordinates ...

Given the introduction of flexitime, even the time keeping is now centrally monitored and controlled.

Hiring and promotions are the responsibilities of the centralised personnel services; the first level of ‘supervision’ involved in these processes is the departmental or branch manager. The supervisor is, in most cases, a ‘leading operative’, and the work of supervision differs little from that of other clerks working in the same section. [100]

What is more, the insecurity of employment associated with continual rationalisation of white-collar work and, in the public sector, with government cuts, affects the middle grades in ways in which it does not those who have already arrived at managerial rank – if only because it means reduced promotion opportunities. For these reasons, it would be absurd to write off all white-collar workers above basic grades as ‘managers’. Many, perhaps most of them, will be torn between placing their hopes on promotion to managerial rank proper (which encourages them to identify ideologically with the new middle class above them), and placing their hopes in trade union action (which encourages the sort of collectivist views traditional in the manual working class). Which they choose will depend upon the extent to which such promotion is a realistic option, the extent to which union action is forceful and effective, and the extent to which there is an attractive pull upon them by actions of other sections of the working class.

Today in 1986 contradictory pulls express themselves in the way in which this section of workers identifies more strongly with its own trade unions than was ever the case before the late 1960s, yet continues in the main to back the Tories and the Alliance parties electorally. Typically, its union members voted to affiliate to the TUC by the mid-1970s, but threw out calls for affiliation to the Labour Party overwhelmingly, even when the majority of union activists would favour this.

Socialists have to be active trade unionists in these grades, while understanding the pressures on them from the new middle class above both in terms of ideology and of promotion prospects. In some ways the situation is analogous with that which socialists faced at the end of the last century when it came to the skilled manual working class. This had living standards that placed it well above the mass of unskilled labourers and tended ideologically to be fairly conservative, acting as a channel by which ruling-class ideas filtered down into the unskilled masses. Yet it was the main section of the class to be organised, through craft unions. Socialists had to be active in such organisations, while continually fighting for unity with the unskilled workers and against the tendency of the skilled to see themselves as ‘labour aristocrats’ with a stake in existing society.

But the situation for socialists active in the middle white-collar layers is easier today in two respects. First, very many of the routine white-collar workers are unionised, and even if objective factors often make it harder for them to commit themselves to activism than the middle grades, in any large-scale struggle, their initiative and enthusiasm does exert a powerful influence. Second, there is very strong organisation among the manual workers, alongside whom many white-collar workers find themselves. This can exercise a strong counter-attraction against the pull of the new middle class influences.


The history of the working class is a history of continual change, as the accumulation of capital leads to the growth of new industries and the contraction of others.

At each stage the class is restructured by the dynamic of the system, with new concentrations of workers being thrown together and old concentrations broken down. So for instance, when Engels wrote the Condition of the Working Class in England in 1844 he was referring overwhelmingly to textile workers. Seventy years later anyone talking about the core of the working class meant the workers in heavy industry who played such a role in Glasgow, Belfast, Sheffield and north-east England. By the late 1930s the locus of growth was shifting again, to the motors and light engineering – and to the West Midlands and north London.

Whenever such a shift takes place, there are always those whose fixed focus on old patterns of industrial organisation makes them unable to understand it in class terms. So many former Chartists could not come to terms with the changes which took place in the British economy from the 1850s to the 1870s and became reconciled to ruling-class politics and accommodated to Gladstonian Liberalism. Nearly a century later, at the peak of the growth of semi-skilled light engineering and motor employment in the late 1950s the view became fashionable that the working class in these industries had become ‘embourgeoisified’.

Today in Britain we are living through another such shift, which is compounded by the effects of recession. Once again we are presented with theories which claim that either the working class or the power of the working class is finished. My aim in this article has been to present some evidence about the real shift which is taking place in the structure of the class and its organisation. Of necessity, this has by and large involved giving an account of the objective structure of the class and of the particular attitudes which are prevalent at a particular point in time. But that has meant extracting a still picture from a moving film. For the consciousness of a class is always in motion, changing much more rapidly than the objective structures of society. Every small victory gives some people new confidence and new understanding; every small defeat spreads some degree of demoralisation, hopelessness and acquiescence in the status quo. There are periods when the victories of the class seem to build on one another, creating a momentum which draws all its weakest and least confident members into a feeling of identity and dragging in its tow sections of other classes; there are periods in which the momentum is lost temporarily and defeats pile up on one another, breaking the sense of class identity even in some of the stronger sections.

In periods of upturn of struggle, consciousness and activity of a class can rise above its strength as measured in cold, objective terms (just think of the Paris Commune!). But conversely, in periods of downturn, consciousness and activity can fall below its objective strength. This is the sort of phase we are going through at present. The demoralisation felt by many activists provides ready soil for the growth of fashionable theories which hold that temporary defeat is permanent decimation.

What I have attempted to do in this article is to show what the objective situation really is. And it is not all that different to that which, in the early 1970s, enabled the organised sections of the manual working class to enter into struggles whose victories brought down a government and, in doing so, demonstrated the value of trade union organisation and industrial action to semi-organised manual workers in sectors like the hospitals and to very wide numbers of middle and lower-level white-collar workers. This was tangible disproof to those ‘theorists’ who had written off the working class in the late fifties.

At some time in the next few years we can expect a change in the level of struggle which will produce just as tangible a disproof of the theories of the Gorzes, the Hobsbawms, the Bassetts, and the Edmondses. The factual evidence suggests that when this occurs, much of the initiative will be taken by ‘traditional’ groups of manual workers, but that they will be followed into struggle by bigger sections of other workers than ever before.


1. See his speech at the TUC Congress, 1 September 1986, and also his interview with Bea Campbell, Marxism Today, September 1986.

2. Philip Bassett, Strike Free: New Industrial Relations in Britain (London 1986).

3. Marxism Today, March 1984, page 9.

4. Marxism Today, October 1982, page 11.

5. See, for instance, John Atkinson in IMS Review, volume 1, Summer 1985, and John Atkinson and Denis Gregory, Marxism Today, April 1986.

6. Manufacturing plus energy and water supplies.

7. Department of Employment Gazette, February 1986. Nigel Harris gives a rather lower figure of four and a half million manufacturing jobs in What to do about London: the strategies of the GLC 1981-86’, in International Socialism 2:31.

8. For an analysis of job losses in the late 1970s which uses such a distinction, see D. Massey and R. Meehan, Anatomy of job loss (London 1982).

9. E. Batstone and S. Gourlay, Unions, unemployment and innovation (Oxford 1986).

10. Economic Trends 1986.

11. Employment Gazette, January 1986.

12. Paul Roots, Collective bargaining: opportunities for a new approach, Warwick Papers in Industrial Relations, number 5, April 1986.

13. Financial Times, 13 March 1986.

14. Batstone and Gourlay, page 53.

15. Analyses of UK Manufacturing local units by employment size, Business Monitor 1975 and 1984.

16. Daniel and Milward, Workplace Industrial Relations in Britain (London 1983), page 218.

17. Batstone and Gourlay, page 72.

18. Batstone and Gourlay, page 82.

19. B. Willey, Union recognition and representation in Engineering, EEF, March 1986.

20. D.F. Wilson, The Dockers, page 312.

21. Classification of Occupations 1980, OPCS.

22. Social Trends 1982, table 4.8.

23. Social Trends 1986, chart 4.11.

24. Heath, Jowell and Curtice, How Britain Votes (Oxford 1985).

25. Goldthorpe and Payne, Trends in intergenerational class membership, Sociology, volume 20, February 1986.

26. Heath, Jowell and Curtice, page 14.

27. Social Trends 1986, table 4.23.

28. Social Trends 1986, chart 10.15; interestingly, the proportion of the managerial-professional groups going to sporting events, including football, is higher than for skilled manual workers and twice as high as for semi-skilled and unskilled manual workers, see table 10.12.

29. Goldthorpe and Payne.

30. Universities Central Council on Admissions, quoted in Social Trends 1986.

31. How Britain Votes, pages 34-38.

32. National Readership Survey, 1984, in Social Trends 1986, chart 10.9; the chart does not provide absolutely precise figures for manual working class readership, since it lumps routine, non-manual and skilled manual workers – social classes C1and C2 – together.

33. British Social Attitudes Survey, 1984, in Social Trends 1986, table 1.9.

34. Financial Times, 28 August 1986.

35. Interview in Marxism Today, September 1986, pages 17-18.

36. A. Bollard, Technology, economic change and small firms, Lloyds Bank Review, January 1983.

37. Atkinson and Gregory, page 13.

38. Atkinson and Gregory, page 14.

39. Batstone and Gourlay, page 8.

40. IDS, Focus, No.39, May 1986.

41. See the tables on rate of unionisation and the use of part-time work in Batstone and Gourlay.

42. Batstone and Gourlay, page 113.

43. Focus, no.39, May 1986.

44. Batstone and Gourlay, page 112.

45. Batstone and Gourlay, page 116.

46. Quoted, Socialist Worker, 23 September 1986, page 13.

47. See E.A. Nordinger, The Working Class Tories (London 1967).

48. D. Lockwood, Blackcoated Worker (London 1958).

49. Crozier, World of Office Work (Chicago 1971) page 17.

50. Lockwood.

51. Crozier, page 16.

52. Crozier, page 18.

53. Crozier, page 17.

54. For example by Goldthorpe and Payne.

55. Goldthorpe, in Giddens and Mackenzie.

56. Social Trends 1982, table 4.8.

57. Classification of Occupations, OPCS, 1980.

58. Crompton and Jones, page 20.

59. Crompton and Jones, page 20.

60. Crompton and Jones, page 186.

61. Crompton and Jones, page 186.

62. Crompton and Jones, page 186.

63. N. Nicholson, G. Ursell and P. Blyton, The Dynamics of White Collar Trade Unionism (London 1981), page 175.

64. Source: G.S. Bain and R. Price, Profiles of Unions Growth.

65. Willey.

66. Heath and others, table 3.2.

67. Social Trends 1986, table 5.2.

68. Crompton and Jones, page 57.

69. Crompton and Jones, page 78.

70. Crozier, pages 33-4.

71. Crozier, page 137.

72. Nicholson, Ursell and Blyton, page 119.

73. See A. Rogers, in International Socialism 2:32, and Jones, page 156.

74. Interestingly, female routine white-collar workers make up the great majority of the readers of Mills and Boon novels, which are almost invariably about upward mobility through marriage – see George Paizis, unpublished PhD thesis.

75. Nicholson, Ursell and Blyton.

76. Batstone and Gourlay; the militancy of the smaller workplaces was probably because they were in the two sections worst hit by increased workloads – the DHSS and the Department of Employment.

77. Labour Force Survey for 1985, in Department of Employment Gazette, May 1986. Heath et al. go as far as to claim that what they call ‘the salariat’ accounts for 30 per cent of men and 23 per cent of women, and has grown from being about a third of the size of the manual class in 1964 to ‘almost the same size’ today (page 35). Goldthorpe and Payne’s ‘service class’ is rather smaller than this – amounting to 18 per cent of men between the ages of 31 and 75, and only about 30 per cent the size of the manual working class on their definitions.

78. For an interesting, although flawed, account of some of these experiences, see B. and J. Ehrenreich, in P. Walker.

79. Nicholson, Ursell and Blyton, page 121.

80. For an attempt to spell out how this works, see Eric Olin Wright, The value controversy and empirical research, in Ian Steedman and others, The Value Controversy (London 1981).

81. Social Trends 1986, chart 5.5. Of course, wage levels alone do not determine class: someone can be relatively highly paid compared to other workers and still be exploited if they are skilled and productive enough. This applies, for instance, to many computer programmers and systems analysts.

82. Social Trends, page 96.

83. Social Trends, page 101.

84. Social Trends, pages 137-8.

85. Goldthorpe and Payne.

86. John Child and Bruce Partridge, Lost managers: supervisors in industry and society (Cambridge 1982), page 3.

87. Crompton and Jones, page 195.

88. Nicholson and others, page 119.

89. Nicholson and others, page 175.

90. Nicholson and others, page 122.

91. Nicholson and others, page 179.

92. Nicholson and others, page 179.

93. Crompton and Jones, page 193.

94. Crompton and Jones, page 156.

95. Nicholson and others, page 179.

96. Nicholson and others, page 179.

97. Nicholson and others, page 179.

98. Nicholson and others, page 107.

99. This was, for instance, the rationale behind the Houghton Report’s attempt to provide for a ‘career structure’ with big differentials between grades in teaching in the mid-1970s.

100. Crompton and Jones, page 65.

Last updated on 29.2.2012