Myths of globalisation and the new economy

Issue: 121

Bill Dunn

Social scientists and journalists have bandied about terms such as “globalisation” and the “new economy” for some time. Behind much of this lies the argument that the working class is dead and with it Marxist hopes of working class self-emancipation. It is tempting to dismiss these claims as just the latest version of anti-Marxist ideology—and, indeed, many of those who now pronounce the end of the working class had little sympathy with it in the first place.

However, many anti-capitalist activists express similar viewpoints and argue for a shift towards new forms and focuses of resistance. Sometimes these recall older anarchist traditions,1 but more recognisably Marxist writers also accept elements of this perspective. David Harvey, for example, thinks economic changes have made earlier forms of left wing organisation “inappropriate”.2 Therefore, while we often have to cut through some fairly crude pro-capitalist propaganda, arguments about restructuring should be taken seriously. Economic structures do not directly determine how we think and act, but central to Marxism is a claim that capitalism creates a collectively exploited working class in a unique position to transform itself and the world. A fundamental restructuring might undermine the bases of action and the prospects for that collective transformation.

There are two distinct sets of claims about capitalism’s change and labour’s weakness. First, geographical shifts are said to have undermined local and national strategies of resistance. This works directly, as capital’s mobility allows it to find ever-cheaper labour and to play workers in one place off against those in another. It also works indirectly, undermining the capacity of nation states to deliver pro-labour and welfare reforms.3 Finance in particular can simply disappear to wherever states provide a conducive environment. Capital’s mobility thus extracts more “corporate welfare” while the burden of paying for it shifts to the poorest and least mobile, especially to workers.4 So globalisation throws workers in different places and countries into ever sharper competition with each other.

Second, it is claimed that a “disorganised capitalism” means widening social differences.5 New technologies, particularly in information and communication, supposedly transform both corporate structures and the nature of work. The giant factories of industrial capitalism are replaced by complex networks of smaller firms, while an increasing emphasis on knowledge and “symbolic manipulation” makes physical labour less important.6 In this “new economy” some skilled workers do well but as individuals not collectively. Other work is degraded and, with more competition for fewer unskilled jobs, those at the bottom suffer. Thus, even according to more sober accounts, there is greater polarisation within the working class, with growth at the two ends of the skill spectrum historically least susceptible to organisation.7

Many academic careers have been made from identifying and labelling these transformations. Meanwhile, a minority of critics insist too much is made of them. There are also many continuities, historical precedents and contemporary qualifications to the picture of change. There is no absolute measure by which to weigh change against continuity and so the academic merry-go-round can continue. However, for Marxists, socialist strategy provides the essential touchstone. If the claims of globalisation and the new economy help to identify new possibilities for struggle or close off others, then these are useful concepts. For many on the left globalisation makes internationalism even more urgent, a “transnational collective response” the only meaningful option.8 Meanwhile, a more diffuse capitalism requires “new strategic imaginations”9 to look beyond the workplace to community and identity based politics or at least to forge links with this to develop a social movement or community unionism.10 This article contests the need for such radical strategic reorientation by evaluating claims of capital mobility and relocation, its impact on state capacity, corporate restructuring and changes in the nature of work.

The limits to capital relocation

For many, globalisation involves a “manic logic” as firms “race to the base” in search of cheap labour.11 Firms go “hopping and skipping and jumping”12 to wherever they can find competitive labour markets, laying waste to jobs in the rich countries. As John Holloway pointed out in an important article in 1995, “capital moves”.13 It is in its nature to move—across borders as well as within them. Unlike in earlier systems such as feudalism, economic activity is not tied to territory. Recent changes are often supposed to have brought a new dimension to this and to the problems it causes labour.14 In the extreme, the economy is reckoned “weightless”. Physical goods are moved more easily as transport technologies improve and, crucially, the economy becomes less physical. Immaterial goods, most notably in finance, move effortlessly around the world. However, this grossly overstates the ease of moving physical goods and the decline of the material economy.15 More fundamentally, capitalism is a social relation not reducible to physical things, so changes to its mobility cannot be deduced from changes in the physical form of commodities.16 This is not to dismiss claims of transformation but to argue that they need to be carefully and critically evaluated.

Historically the extent of capital relocation has varied enormously. There was an earlier phase of internationalisation in the 19th and early 20th centuries. By 1914 outward foreign direct investment (FDI) from Britain amounted to 53 percent of GDP, a level comparable to today’s.17 Levels from other rich countries were lower but still not exceeded until the 1980s or 1990s. Many familiar multinationals were already well established. There was much less FDI in the inter-war period. During the long post-war boom there was a still greater concentration of investment within rich countries’ domestic economies. What FDI there was also went primarily to other rich countries. So in 1914 half, and in 1938 two thirds, of the total had gone to poorer countries but that fell to just 20 percent by 1960.18 Capital does seek cheap labour, but there are many political and economic reasons why investment can profitably be concentrated and why relocation can prove difficult.

Of course, just its potential to move gives capital power.19 Even in the 1980s, before talk of globalisation became fashionable, managers in the US car industry successfully “whipsawed” plants in different places, demanding concessions on wages and conditions. However, employment in the car industry in rich countries actually increased between 1970 and 2001.20 Firms, and capital’s political supporters, talk up mobility to threaten workers or to extract bribes from governments. For instance, the United Nations published a survey in 2005 showing that companies claimed China was their overwhelming favourite prospective overseas business location. India ranked second and four other poorer countries were in the top ten: Russia fourth, Brazil fifth, Mexico sixth and Thailand ninth.21 But in practice, the following year 66 percent of overseas investment went to established rich countries and only China (fifth) and Hong Kong (seventh) were among the top ten destinations. Russia was 11th, Mexico 18th, Brazil 19th, India 21st and Thailand 27th.22 Looking at the extent to which capital actually moves seems a necessary starting point for assessing claims to its mobility.

Real structural changes do give substance to recent claims of globalisation. World trade jumped from $519 billion to nearly $12 trillion between 1973 and 2007, and within this the share of what are classified as “developing countries” rose from 24 percent to 39 percent.23 Similarly, stocks of FDI increased from $560 billion to $12 trillion between 1980 and 2006, or from 5.3 percent to almost 25 percent of world GDP. The poorer country share went back over 30 percent.24 Along with, but not reducible to, foreign investment, the level of manufacturing in developing countries shot up. It rose by 75 percent and from 19 percent to 30 percent of the world total between 1990 and 2005.25 However, these aggregates have to be interpreted quite cautiously.

First, as of 2004, 71 percent of the world’s industrial production was performed in the 21 richest countries, with GDP per capita greater than $20,000, which between them had only 14 percent of the world’s population.26

Second, changing patterns of production were hugely uneven. Just five countries, China and the “Asian tigers”, accounted for 62 percent of the poorer country growth from 1980 to 2005.27 FDI also went to a few favoured locations, mainly in Asia, which were seldom those with the lowest wages. By 2006 just five countries—Hong Kong, China, Mexico, Brazil and Singapore—accounted for well over half the developing world’s total FDI. Africa had received less than 1 percent.28 There was no “race to the base”. The rise of industry in poor countries (and its relative decline in many rich ones) also often reflected changes within these countries as they got richer, rather than simply a movement of production from the rich to the poor. If productivity grows quicker in manufacturing than services (as it normally does), then even if consumption patterns stand still, manufacturing’s share of employment shrinks. Indeed, industrial employment was falling in China, even before the 2008 economic crisis hit.29 In terms of the origins of investments, firms from a few rich countries predominated. This seems important in relation to arguments of deindustrialisation. The UN’s figures, although almost certainly exaggerating the picture, indicate significant net outflows, amounting to $2 trillion, away from the richest countries during the 1990s, particularly from Britain, France and Germany, where they averaged 3.9, 2.4 and 1.6 percent of GDP respectively. These are significant sums, and plausibly contributed to slow growth and deindustrialisation. However, in the US where deindustrialisation was probably sharpest, net outflows amounted to only 0.3 percent of GDP per year during the 1990s. In the early years of the 21st century there was then a modest retrenchment and net inward investments across the rich countries.30

Third, only about 30 percent of FDI was in manufacturing, and only 8 percent in poorer country manufacturing. Of this only a small proportion was wholly new investment. In 2006 67 percent of all FDI involved buying existing assets and 30 percent involved reinvesting the earnings of the already operating foreign plants. So less than 3 percent was new or “greenfield” investment.31 But even taking what are therefore vastly exaggerated aggregate figures, the largest poor country recipient, China, received net FDI inflows amounting to 9.2 percent of the level of fixed capital formation and 3.2 percent of GDP in 2006.32

This highlights, fourth, that poorer country growth does not necessarily depend on rich country investments. Among other things, multinationals also subcontract to local suppliers, notoriously in the case of sweatshops in the textile industry. This could have a similar effect on rich countries. The US and Britain both recorded big trade deficits, which have been interpreted as meaning a loss of jobs. However, most analyses suggest that trade, and trade deficits, account for at most a small proportion of unemployment in rich countries.33 In the US unemployment of just over 7 million in 2006 was nevertheless lower, in both relative and absolute terms, than in 1980 or 1990, when trade and trade deficits were a much smaller proportion of GDP.34 Elsewhere, notably in Germany and Japan, there were persistently big trade surpluses but workers experienced many similar problems.

Fifth and finally, there was huge unevenness between industries. In some sectors such as textiles, clothing and shoes, toys and consumer electronics something of a rush of manufacturing to poor countries did happen. Here claims of globalisation as a cause of labour’s problems may seem appropriate. The clothing industry provided the classic case for arguments of a new international division of labour35 and this has been confirmed in more recent examples of relocations and mass lay-offs in rich countries. Levi Strauss, for example, from a peak of 28,000 employees had, by 2003, ended all manufacturing in the US.36 Again this was hardly “global” and 45 percent of textile exports were accounted for by just seven Asian countries.37 And even in clothing there were significant exceptions. Some firms, the Spanish company Zara is the best known example, retained concentrated production systems within their home country. Italy and Germany, concentrating on upmarket lines, remained second and third to China among textile exporters.38 A similar logic seemed to drive consumer electronics. So by 2006 Mexico and China accounted for 21 percent and 17 percent of the world’s exports of TVs, distantly followed by Turkey, Japan, Poland and the Netherlands (all selling around 4 percent of the total).39 But here too, rather than racing to the base, there was a rationale by which production was re-concentrated. Japanese firms invested in eastern China and stayed even as wages there rose.40 Therefore, even in these sectors there are important qualifications. Moreover, these are important, and particularly visible, sectors but not necessarily typical.41

Other industries did not go offshore to anything like the same extent. Two thirds of the world’s automobiles were made in just seven countries: Japan, Germany, France, the US, Korea, Spain and China.42 And while the last of these became a major producer, it remained a substantial net importer. Rich countries dominated export markets.43 The car industry is the classic example of capital winning bribes from governments, including state governments in the US, to locate within their borders, sometimes at truly exorbitant cost.44 Nevertheless, the net effect was that even some of the most well known “global” manufacturers were more national than global in terms of their assets, sales and employment. General Motors had nearly two thirds of its workforce and assets in the US, Toyota two thirds of its workforce and half its assets in Japan.45 Again, particularly for US and European firms, the foreign assets and employment were also overwhelmingly in other rich countries. A much higher proportion of overseas Japanese auto investments were in poorer countries, 49 percent of the total going elsewhere in Asia.46 Japan nevertheless remained a substantial net exporter of cars. For most firms a large proportion of the overseas production was aimed at local or sometimes regional markets. Improved transport systems notwithstanding, it can pay to produce cars near their intended market. Parts can be made further away in extended networks, although here too things like “just in time” (JIT) production systems could militate against having suppliers too far away and the production of at least some components became reconcentrated close to the assemblers.

Other goods are very light in relation to their value and can potentially be moved easily. Semiconductor chips are an obvious example, and microelectronics is accordingly seen as an industry at the leading edge of globalisation.47 Unlike in car making, there is little history of labour organisation. However, in terms of industrial structure there are parallels and here too there are significant agglomeration effects. Despite the products’ mobility, shifts in manufacturing location remain concentrated within a few relatively rich countries, albeit these include relatively newly industrialised Korea and Taiwan.48

Tellingly, even in a sector such as finance, whose “products” are in theory almost completely mobile, the workforce remains highly concentrated in rich countries. This applies not only to the city slickers in New York and the City of London but also to routine “production workers”. Off-shoring is a tax dodge, which in finance has little to do with the work of the industry. Most of the tax havens have very low populations and are not a plausible alternative to employment in rich countries. Some firms have relocated some processing and call centres, for example to India, but although well publicised, these represent a tiny fraction of the sector’s employment.

Many other industries are inherently immobile. This is more or less true of many “in person” services—you cannot offshore a haircut for instance—but also of much material production. Construction is the obvious example here. If such workers cannot be threatened by globalisation, others like transport and communications workers might be expected to be positively empowered. One well known advocate of a highly globalised worldview, Susan Strange, admitted exceptions to labour’s weakness among groups such as lorry drivers.49 There are significant examples of militancy, for example the 1997 UPS strike in America and more recently on London underground. But overall the evidence suggests that transport workers are neither more militant nor better rewarded than others.50 It is not obvious that capital’s mobility or immobility has had much impact.

Finally, it is worth noting that although the arguments about capital relocation are often assumed to be generally valid they explicitly account for the weakening of labour only in rich countries. At the very least, one might expect a corollary that a powerful working class has emerged in places where it did not exist before. It has already produced effective labour organising (as well as the familiar problems of organising) in Brazil, South Korea and South Africa, for example, where only a few decades ago unions did not exist.51 It has also created a vast potential in many other places, notably China.

In short, the evidence of capitalism’s globalisation is patchy at best, and its association with workers’ experience even weaker. Capital can flee—and has fled—high wages and labour militancy but to a much lesser extent than its supporters suggest. The implication is that the potential for resistance “in situ” remains considerable.

State retreat and the “democratic deficit”

For many writers, labour’s decline is an indirect result of declining state powers. Reforms won at the national level are eroded as capital mobility overwhelms states. There emerges a “democratic deficit” and a “hollowing out” of welfare functions.52 The implication is that labour should not seek power in an arena from which it has vanished but should upscale to the global level. There is a reformist version of this argument, which looks to global institutions such as the International Labour Organisation or (even more optimistically) the World Trade Organisation, or which sees the international trade union bureaucracy as likely agents of change.53 There are also more radical interpretations that suggest it is necessary to rediscover, but this time with a deeper meaning, traditions of rank and file labour internationalism.54

It is probably not necessary to dwell too long on this perspective here. The discussion in the preceding section already dealt with it, at least implicitly. If capital is not actually as mobile as is often assumed, there are many reasons to think that states could still resist. This argument has also been the focus of a substantial scholarly literature with opponents raising important theoretical, historical and contemporary empirical reasons for regarding state retreat as something of a myth.55 There can be an unfortunate nationalism to these “sceptical” writings but they do successfully highlight just how interventionist even supposedly liberal states actually are.56 In addition, the claims of state retreat misunderstand the nature of the state under capitalism in a way that most readers of this journal will readily appreciate. There can sometimes be important tensions between the state and capital but they are not essentially opposed. It is a reformist illusion that socialists could ever simply capture the state and use it for their own purposes without fundamentally challenging global capitalism. However, state power is not simply reducible to the interests of capital and even a limited autonomy can be important, making it possible to win some reforms at the national level, however inherently insecure they might be. Identifying the limits of reformism never meant rejecting the struggle for reforms and it is possible that the capacity to win and sustain them has now been eroded.

Against this it seems worth making two rather simple points. First, there is little evidence of the state’s economic decline. Levels of state spending, including social spending, increased in most rich countries after 1980.57 Levels contrasts sharply with those before the Second World War, let alone in the 19th century. Nor have high levels of state spending proved an economic disaster, in a way that might indicate their fragility.58 In large rich countries at least, states have access to the same technologies that are supposed to be undermining them, technologies they have little difficulty applying when it suits, for example monitoring funds to Cuba or alleged terrorists.59 Even relatively poor countries like Chile and Malaysia have imposed capital controls rather effectively.

Second, the national level remains a highly contested arena, which labour’s opponents show no sign of abandoning. Many of the attacks on labour have been the direct result of state policy, for example cutting minimum wages, notably in the US, and introducing anti-union laws. The “business community” remains well organised, particularly, if not exclusively, at the national level60 and states respond to the anti-labour agendas of their domestic capitals. Their support for such corporations typically outweighs their desire to attract or maintain external investment.61 In addition, “reforms” are still more or less effectively contested by local labour movements. The resistance to anti-labour laws in France, for example, contrasts with the situation in Britain and elsewhere.

The repeated attempt to impose anti-labour policy shows that mere economic forces are insufficient to discipline labour. The national level remains a vital arena of struggle, which supporters of labour internationalism should be wary of abandoning. In practice it is clear that national labour movements still matter.

Labour and the new economy

There are a mass of claims of a new era of disorganised, flexible, post_Fordist capitalism.62 Above all, this is supposed to have reinvigorated the economy. There are many reasons to doubt this but the broad economic claims cannot be discussed in detail here. Suffice it to say that, although there were real rises in productivity in the US from the mid-1990s, these did not return to levels seen in the 1970s, let alone those of the long post-war boom. Elsewhere, in Japan and continental Europe, similar technological innovation coincided with falling productivity gains and extremely sluggish overall economic performance. However, labour’s decline was common across most rich countries and the focus here will remain on claims that the new economy transforms the prospects for labour organising. There is a certain concentration on the US in the next two sections, where the claims of transformation are perhaps strongest and the decline of manufacturing sharpest, but similar phenomena have occurred and similar causes have been suggested elsewhere. This section considers arguments of capitalist reorganisation and declining workplace size. The next discusses claims of a transformation of labour, achieved as new technologies supposedly change the skill bases of work.

To start with, then, the new economy is perceived to have altered company and workplace structure. Giant “Fordist” factories are replaced by complex subcontracting networks. While claims of a reinvigorated cottage industry63 might seem fanciful, physically smaller units have historically been harder to organise.64 It is also claimed that there is less hierarchy at work. Small capitalists exploit but are themselves victims of the big, and power is diffused across networks, putting class antagonisms out of sight.65

There is clearly a danger of too mechanical a reading of relations between industrial structure and labour’s organisation. It took 38 years of sometimes bloody struggle to organise at the Ford Motor Company itself, while many giant factories today, for example in China, effectively lack independent unions. Conversely, well organised, even militant, groups of workers could be employed in relatively small workplaces. The average pit size in British mining was about 300 at the beginning of the 19th century and only around 900 at the beginning of the great strike of 1984.66 Nevertheless, the way the concentration and centralisation of capital bring together ever more workers has been widely perceived as an important part of the process by which capitalism creates its own gravediggers, and if this is no longer happening it would appear to have serious consequences.

However, concentration and centralisation were only ever tendencies within capitalism. Against this Marx also argued:

If the system of manufacture seizes upon a trade which was previously carried on in connection with others, either as a chief or a subordinate trade, and by one producer, these trades immediately break their connection and assert their independence of each other. If it seizes upon a particular stage in the production of a commodity, the other stages of its production become converted into as many independent trades.67

This breaking up of once connected processes thus reflects the growing, not declining, scale of production. So early semiconductor companies made a range of different microchips and performed all the stages of production themselves, from design through to assembly and testing, and often also their inclusion in finished products such as computers. By specialising, for example in making microprocessors or memory chips or even chips for specific applications such as watches, a few firms could dominate their respective markets. This was most obvious in the case of Intel in microprocessors. Similarly, rather than making their own manufacturing equipment, the chip makers could buy this from a smaller number of specialist firms like Applied Materials, Nikon or Canon, which achieved greater economies of scale than if all the semiconductor firms produced their own capital goods.68

This is not confined to “new economy” sectors. For example, in auto-assembly, with Western firms imitating what was already practised in Japan, an increasing share of parts production has been outsourced to specialist component makers. This means the apparent diffusion of production actually often represents growth beyond what was contained within a single workplace.69 This has significant implications. It means that workers in different parts of the network can be just as tightly connected as those within a single factory. A strike in a small parts supplier can disrupt the whole system. Andrew Herod’s account of the 1998 strike at a single General Motors component plant in Michigan demonstrates how this could quickly disable production across North America, costing the company over $2 billion.70

Sometimes this differentiation of particular tasks and stages of production can occur within a single firm but the particular units can then also be set in competition with each other or spun-off as independent suppliers. The development of such networks has allowed the final producers to increase their sales without necessarily increasing their employment as they have subcontracted larger parts of their production to independent suppliers. This can complicate class relations so the immediate employer and the end contractor pass on responsibility for cost cutting and attacks on labour. However, the relations of independence are often more formal than real. Sometimes, as for example in the UK construction industry, many nominally small firms are actually single employees redefined as a tax dodge.71 Even between real firms, nominally free markets often conceal effective power relations between the tiers of suppliers.72 With more than one supplier of any particular component, this could be an effective strategy for setting the rival firms and their workers against each other. However, it also implied similar possibilities for dislocating the whole network through action in any of the parts. One of the features of such networks has been the tightness of just in time production schedules so the other firms cannot simply step in if something goes missing elsewhere. For example, 230 strikers at an independent Australian bumper and dashboard maker in 2007 could apparently affect Ford’s assembly within 40 minutes and cost it
$1 million a day.73 Competitors might eventually be able to ramp up production, so without winning solidarity across the different suppliers, action in any one could be undermined in the medium term. But this does not seem to be a qualitatively new problem.

Nor do these processes abolish the tendency towards concentration and centralisation. Again, for the car industry in North America, there was a rapid decline in the number of suppliers from the 1990s as the extent of subcontracting increased—from 30,000 in 1990 to well below 10,000 after 2000.74 That is to say, their average size grew particularly quickly, many becoming industrial giants in their own right. Perhaps the most telling fact is that union density held up much better in the assemblers, where firm size fell, than in the parts sector, where the firms got bigger.

A similar picture emerges in other sectors. In finance, American workers have hardly been organised since the United Post Office and Professional Workers of America was witch_hunted from the CIO in 1950.75 In the UK and many other European countries bank workers established quite effective unions despite their relatively small workplaces. However, significant restructuring saw a retreat from branch banking and a greater concentration within central processing and call centres. These new larger concentrations were initially unorganised. Only later, for example among UK banks, were there significant victories in terms of recognition and recruitment.76

The most fundamental point seems to be that the image of a return to small-scale production is at best hugely overstated. For the US average workplace size simply did not fall, and the proportion of people working in units of 500 or more rose from 46 percent to 49 percent between 1990 and 2004.77 There seems little reason to believe that the picture is radically different in other rich countries, while in many poorer ones vast new concentrations of labour were created. The process of outsourcing is also often associated with the off-shoring of production discussed above. It was suggested this practice was far from universal. However, it is very real in some sectors with the classic image of shoe or clothing companies subcontracting to tiny sweatshops in East Asia. The conditions for workers are undoubtedly often grim. But this is hardly the result of the small scale of operations. To give just one example, in 2005 in Cambodia’s export garment industry, only 28.6 percent of the factories had under 500 employees and 7.3 percent had over 5,000. For most of the sector’s 260,000 workers, large workplaces were the norm.78

Continuities, even increases, in average plant size might, of course, mask real changes. Many formerly big workplaces declined, undermining hard won union strengths. Meanwhile, new concentrations of labour did not produce organisation without struggles. The point is simply that this organisation and these struggles are not precluded by structural change.

The changing nature of work

The second set of claims for labour’s reduced potential invokes changes in the nature of work. In particular, new technologies supposedly change the skills needed. This creates a polarising process. Some jobs need more skills. This is most obviously the case for the key “symbolic manipulators” in research and development or in finance, so central to many characterisations of the new, knowledge economy.79 However, methods of teamwork among manufacturing workers also mean reskilling and according to some writers even create a “neo-artisanate”,80 working more flexibly, less alienated in their labour and less likely to seek collective solutions.81 Meanwhile, the new technologies deskill other jobs. The complex skills of the bank teller and even the calculations of the bar worker can be replaced by the simplest of computers. Some low skilled, low-tech jobs such as cleaning remain relatively untouched but with fiercer competition for fewer such jobs wages are driven down. These jobs are casualised, often part time and flexible—but the flexibility here is all that of capital against labour.82 This means a decline in the classic semiskilled working class, once the core of labour organisation. According to this view, it is the skill polarisation that lies at the root of the widely observed increases in patterns of inequality. This process is then also often perceived to accentuate labour market inequalities based on gender and ethnicity.83

Again, however, there are both theoretical and empirical reasons to be very cautious about these arguments. As Lawrence Mishel and his colleagues argue, “it is generally true that investment and technological change are associated with the need for more workforce skill and education—but this was true for the entire 20th century, and it therefore does not explain why wage inequality began to grow two decades ago”.84 Nor is it obvious why the semi-skilled should be particularly prone to organising. Historically both high skilled (teachers, for example) and low skilled (rubbish collectors, for example) have been well organised, even militant.

Often it is hard to evaluate the effects of technology on skills. In some high-tech industries such as micro-electronics there may well have been a polarisation between skilled technicians and routine production workers.85 Elsewhere the opposite happened. The construction industry offers one relatively clear example. Here new technologies compressed skill requirements towards the centre. They reduced the relative importance of skilled on-site craftwork but also reduced the need for manual lift and carry work.86 There is little evidence that this helped labour organisation, which in most countries did no better than in other sectors and in the US did markedly worse.

There seems a real danger of conservative circularity in arguments about skill, pay and organisation. Mainstream economics simply asserts that people are rewarded for their “human capital”, for their skills. (Since their wages are lower we should also presumably assume that women and ethnic minority workers are systematically less skilled than white men.) But if we then look at the distribution of incomes, for example in the US, it is true that both the top and bottom fifths are less well unionised than the middle three.87 Of course, for Marxists, the line of causation might run in the opposite direction (at least at the bottom). People are low paid because they are badly organised, not the reverse. We know unions win wage rises so it is not surprising to find those on the lowest wages less well organised than the middle.88 Indeed, in the US, the growth of within-group inequality, that is among those with similar education, accounted for about 60 percent of total growth of wage inequality from 1973 to 2005.89

Historically, many low-skilled traditionally male jobs were once casualised and badly paid until they were effectively organised. The same holds true today, so that women and non-white workers are disproportionately in low-paid jobs. But the “union premium” is all the greater. In the US the pay advantages of unionising are 24 percent for men but 31 percent for women. They are 30 percent for white workers, 32 percent for black workers and 46 percent for Latino or Hispanic workers.90 Unsurprisingly, studies report that women and ethnic minority workers tend to have more positive attitudes towards unions than white men.91 There are still big gender and racial differences and all sorts of practical and institutional obstacles to organisation and to winning equal pay but the idea that their different work or human capital makes labour organisation inappropriate is nonsense. Instead the erosion of unionisation accounts for most of any increasing wage polarisation.92

There were changes in many production processes but any association with labour’s disorganisation should also be treated cautiously. Inspired by work practices introduced in Japan, particularly at Toyota (after severe defeats for labour in the late 1940s and early 1950s), ideas of teamworking were introduced in many Western industries. Rather than a rigid job demarcation characteristic of Fordist production lines, work groups took responsibility for a wider range of tasks. In Japan this was also associated with “lifetime” employment, relatively secure jobs at least for core, mainly male, workers until their 50s. Firms supposedly invested more in their workers’ skills and workers had more rewarding jobs. Systems of “continuous improvement” valued workers’ input into the production process, which itself was demand driven, with goods only produced as they were needed rather than pushed through by the production line.93

Where such systems have been implemented they may have brought a degree of re-skilling. But there is little evidence that it leads to a systematically less alienated working environment. It was more a way of making people work harder. Where the rigidities of the production line left moments of downtime as not all jobs could be performed in simple multiples of each other, flexibility meant filling in such gaps. The next job was always waiting.94 Similarly, teamworking meant someone could always fill in to cover if someone else was away, off sick or having a break. Thus it produced what has been termed “management by stress”.95

The rise of service sector jobs does not fundamentally change things. Services are susceptible to similar processes to manufacturing but also to labour organisation. The degradation of white collar work identified by Harry Braverman’s classic study96 has often been intensified. Even among very highly skilled “symbolic manipulators” such as computer programmers, the sheer size and complexity of many processes mean that any one programmer may only write a tiny part of the whole, with workers sometimes doing this in vast factory-like concentrations. Meanwhile, it is perhaps also in services that management by stress becomes clearest. In call centres an ability to handle a range of enquiries, in however rudimentary a manner, guarantees that a pool of workers always have a call waiting. At Heathrow Airport uniting the check-ins, combining all the queues, guarantees horrible waits for the passengers but also ensures the staff never have a moment without an angry customer to serve as quickly as possible. This all makes working life grim and squeezes more value out of labour. However, it does not preclude resistance.

Braverman’s book was criticised for too determinist a reading of the way change is imposed on workers. Both blue and white collar workers can more or less effectively resist and shape their working environment. And things like “Japanese” practices could turn out differently when introduced to Western workplaces with different labour traditions. Team meetings could become forums for voicing grievances, not for improving productivity—and in many cases were quickly abandoned. Competition between teams could mean sabotaging the other, rather than intensifying the work effort.97 In fact, these systems could become more vulnerable to disruption, not less, and they have tended to need more supervision to make them run smoothly.98 Employers have often recognised the need to negotiate with unions to achieve this, and unions recognised their ability to cut deals to limit exploitation.99

So flexibility was more a strategy against labour than a reflection of a fundamentally new economic logic. It could involve a redefinition of jobs and breaking contracts rather than a transformation of work. Even this, however, was often more an ambition for capital rather than an achievement. Between 1995 and 2005 there was actually a slight decrease in the proportion of the US workforce doing “non-standard” work, while even among non-standard workers an increasing proportion (up from 62 percent to 65 percent) were working as “permatemps”—they had casual contracts but stayed in the same job for more than a year.100

Finally, and perhaps most significantly, the simple fact is that class polarisation among workers has not really happened. Lawrence Mishel and his coauthors argue that the greatest wedge has been between the top 10 percent and everybody else101 and claims of polarisation are usually based on this comparison. This is hugely misleading as a measure of dispersion within the working class. Of course, executive pay has spiralled and a few senior managers done very well. Comparing instead the lowest incomes with the median, those in the middle of the pay scale, there is a much more ambiguous picture. Here the gap “grew in the 1980s but has been stable or declining ever since”.102 In particular since 1987 (just as the new economy was supposed to be kicking in) the difference declined for men and remained roughly stable for women.103 So for women in the US, there was a clearer across the board polarisation than for men, with significant increases in pay for the top 80 percent. However, this reflected a narrowing, not widening, of gender inequality. So although in 2005 29.4 percent of women but only 19.9 percent of men still earned what were officially admitted as poverty wages, for men this was up from 15.7 percent and for women down from 42.1 percent in 1979. There was also a decline of poverty wages among black men and women and among Hispanic women but not men.104 In continental Europe too the pay gap between the bottom 10 percent and the mean tended to decline.105

So polarisation within the working class was, at least, much less than is usually claimed. It occurred among people with similar skills and had little to do with any particular technological transformation. Labour organisations’ decline is better understood as the cause of rising inequality and low wages not as the result.

Organising and the failures of organising

Capitalism continually brings workers together and pushes them apart. It does this both literally by creating new concentrations of labour in particular places and setting workers in different places against each other. It also does so figuratively, creating common class interests and identities but also manufacturing a thousand strategies of divide and rule. Socialist politics has always been about overcoming real and perceived differences to win solidarity.106 Recent failures may be at least as much about the failures of left politics as about any objectively increasing spatial or economic heterogeneity. Indeed, Richard Walker reasonably argued that we should understand the contemporary period more in terms of labour’s political defeats than capital’s economic successes.107

Some contemporary writing, for example that which advocates a social movement or community unionism, might be interpreted as an attempt to revive lost traditions of solidarity. However, there is a danger of posing these strategies as an alternative to workplace based organising, and more or less explicitly endorsing the view that there is no longer anything special about workers’ collective exploitation at work that tends to challenge their fragmentation. Socialism becomes simply an act of will. The real Marxist tradition has always been internationalist and sought to organise beyond the workplace—to become, in Lenin’s phrase, a “tribune of the people” rather than simply a trade union secretary.108 But it also argued that there was something special about exploitation and resistance at work, which created labour’s unique potential to transform the world.

Capitalism, as Marx and Engels long ago insisted, continually changes. However, there seems little reason to believe that recent shifts bring a radically new dimension to some rather old problems for workers and socialist organisation. Even in manufacturing, capital flight is rare. Capital cultivates the idea that it can easily escape from militancy and high wages but it can seldom move as frictionlessly as it purports. Huge wage differentials did not produce a rush of investment from rich countries to poor. There were significant movements, but in specific sectors and often then in highly uneven ways. In many sectors of capital, in declining primary industries but also in many expanding service sectors, there was still less prospect of capital movement. Labour’s problems were at most weakly associated with capital’s mobility. Similarly, although corporate restructuring did undermine many previously established labour strengths, it produced neither a return to small-scale cottage industry nor the end of alienation at work. It did not preclude effective workplace based strategies for organising. None of this is to dismiss the importance of scale nor thinking tactically how disputes can best be spread and strengthened. Circumstances vary. If a European shoe factory is being closed down and production shifted offshore, any chance of keeping it open may indeed depend on forging external links. It would be politically stupid to dig up figures showing that its experience was actually atypical. But it would be equally misguided to argue that all workers faced the same problems. Rail workers, construction workers, hospital workers, for example, and people in a raft of other jobs including manufacturing jobs—while of course well advised to broaden any sectional struggles—have real power at work. The possibility of fighting immediate exploitation at work provides an enduringly essential starting point for socialist internationalism.


1: For instanace, Hardt and Negri, 2000.

2: Harvey, 2003.

3: Rowley and Benson, 2000.

4: Frieden, 1991; Strange, 1996.

5: Castells, 1996, 1997 and 2000; Lash and Urry, 1987 and 1994; Murray, 1988; Piore and Sabel, 1984; Hyman, 1992; Reich, 1991.

6: Reich, 1991.

7: Hyman, 1999.

8: Tilly, 1995; Mazur, 2000; Radice, 1999.

9: Hyman, 1999.

10: Waterman, 1999; Wills, 2001; Wills and Simms, 2003.

11: Greider, 1997.

12: A Glassman, cited in Reich, 1991, p121.

13: Holloway, 1995.

14: Ross, 2000.

15: Huws, 1999.

16: Holloway, 1994 and 1995; Fine, 2004; Dunn, 2004.

17: Held, McGrew, Goldblatt and Perraton, 1999, p275.

18: Kenwood and Loughheed, 1992; Dunning, 1993.

19: Thomas, 1997.

20: Pilat, Cimper, Olsen and Webb, 2006.

21: UNCTAD prospects assessments 2005, available at

22: Calculated from UNCTAD World Investment Report 2007, available at

23: WTO “International Trade Statistics” for 2007, available at Their definition of developing counties is a static one, which includes the now relatively wealthy countries in East Asia. However, it does have the advantage of allowing comparisons of a constant group of countries.

24: UNCTAD, World Investment Report 2006, available at

25: UNCTAD, Handbook of Statistics 2006-7.

26: My calculations, based on figures in United Nations Development Programme Human Development Report 2006; United Nations Commission on Trade and Development World Investment Report 2006; World Bank, World Development Report 2006: Equity and Development; Census, Statistical Abstract of the United States 2008, available at

27: World Bank, World Development Report 2006: Equity and Development.

28: UNCTAD, World Investment Report 2007.

29: UNCTAD, Handbook of Statistics 2006-7; National Bureau of Statistics of China, available at

30: UNCTAD, World Investment Report 2007. The exaggeration is apparent from the global figures, which suggest the world experienced a net outflow of over $1 trillion during the 1990s. Globally, net flows should, of course, be zero.

31: UNCTAD, World Investment Report 2007.

32: UNCTAD, FDI country profiles, 2007, available at

33: Nayyar, 2007); Rowthorn and Ramaswamy, 1997; Navarro, 2000.

34: Census, Statistical Abstract of the United States 2008, available at

35: Fröbel, Heinrichs and Kreye, 1980.

36: Dicken, 2007.

37: Comtrade, available at

38: Comtrade.

39: Comtrade.

40: Thun, 2008.

41: Sutcliffe and Glyn, 1999.

42: Dicken, 2007.

43: Comtrade.

44: Dicken, 2007.

45: Dicken, 2007, p296.

46: UNCTAD, FDI country profiles, 2005, United States, Japan, available at

47: Henderson, 1989.

48: Dicken, 2007, p333.

49: Strange, 1996.

50: My calculations from the ILO International Statistics Yearbook for various years, see Dunn, 2004.

51: Moody, 1997.

52: Burnham, 1997. Burnham, it should be stressed, criticises this view.

53: Boswell and Stevis, 1997; Mazur, 2000; O’Brien, 2000; Hughes, 2002.

54: Waterman, 1999; Tilly, 1995; Radice, 1999.

55: Hirst and Thompson, 1999; Weiss, 1999.

56: Harman, 1996; MacLean, 2000.

57: Glyn, 2006.

58: Garrett, 2000.

59: Henwood, 1998; Gowan, 1999.

60: Henwood, 2006.

61: Herod, 1991.

62: See references in note 5.

63: Sabel, 1984.

64: Ackers, Smith and Smith, 1996.

65: Lash and Urry, 1987.

66: Benson, 1980; and calculated from Callinicos and Simons, 1985, who give figures of 184,000 working in 198 pits.

67: Marx, 1976, pp473-474.

68: Chon, 1997

69: Moody, 1997.

70: Herod, 2000.

71: Harvey, 2001.

72: Gereffi, Humphrey and Sturgeon, 2005.

73: ABC News, Australia, 24 August 2007.

74: Dicken, 2007, p292.

75: Pollard, 2005.

76: Bain and Taylor, 2002.

77: Census, Statistical Abstract of the United States, available at

78: ILO, 2005.

79: Reich, 1991.

80: Piore and Sabel, 1984.

81: Womack, Jones and Roos, 1990.

82: Pollert, 1988.

83: MacEwan and Tabb, 1989; Cox, 1996.

84: Mishel, Bernstein and Allegretto, 2007, p199.

85: Dicken, 2007.

86: Thieblot, 2002.

87: Mishel, Bernstein and Allegretto, 2007.

88: Mishel, Bernstein and Allegretto, 2007, p188.

89: Mishel, Bernstein and Allegretto, 2007, p200.

90: Calculated from Census, Statistical Abstract of the United States, available at

91: Bronfenbrenner, 2003.

92: Mishel, Bernstein and Allegretto, 2007, p7.

93: Womack, Jones and Roos, 1990.

94: Smith, 2000.

95: Parker and Slaughter, 1988.

96: Braverman, 1974.

97: Clarke, 1997; Rinehart, Huxley and Robertson, 1997; Walton, 1997.

98: Delbridge and Lowe, 1997; Lewchuk and Robertson, 1997; Murakami, 1997; Rinehart, 1999.

99: Thelen and Kume, 1999.

100: Mishel, Bernstein and Allegretto, 2007, pp238, 241.

101: Mishel, Bernstein and Allegretto, 2007, p210.

102: Mishel, Bernstein and Allegretto, 2007, p5.

103: Mishel, Bernstein and Allegretto, 2007, pp142, 201.

104: Mishel, Bernstein and Allegretto, 2007, pp124-127.

105: Glyn, 2006.

106: Panitch, 2001.

107: Walker, 1999.

108: Lenin, 1975, p99.


Ackers, Peter, Chris Smith and Paul Smith, 1996, “Against All Odds? British Trade Unions in the New Workplace”, in Peter Ackers, Chris Smith and Paul Smith (eds), The New Workplace and Trade Unionism (Routledge).

Bain, Peter, and Phil Taylor, 2002, “Ringing the Changes? Union Recognition and Organisation in Call Centres in the UK Finance Sector”, Industrial Relations Journal, volume 33, number 3.

Benson, John, 1980, British Miners in the Nineteenth Century: A Social History (Gill & Macmillan).

Boswell, Terry, and Dimitris Stevis, 1997, “Globalization and International Labor Organizing: A World-System Perspective”, Work and Occupations, volume 24, number 3.

Braverman, Harry, 1974, Labour and Monopoly Capital (Monthly Review).

Bronfenbrenner, Kate, “The American Labour Movement and the Resurgence in Union Organising”, in Peter Fairbrother and Charlotte Yates (eds), Trade Unions in Renewal: A Comparative Study (Continuum).

Burnham, Peter, 1997, “Globalisation: States, Markets and Class Relations”, Historical Materialism 1.

Callinicos, Alex, and Mike Simons, 1985, The Great Strike (Socialist Worker).

Castells, Manuel, 1996, 1997 and 2000, The Information Age: Economy, Society and Culture, volumes 1, 2 and 3 (Blackwell).

Chon, Soohyun, 1997, “Destroying the Myth of Vertical Integration in the Japanese Electronics Industry”, Regional Studies, volume 31, number 1.

Clarke, Louise, 1997, “Changing Work Systems, Changing Social Relations?”, Industrial Relations, volume 52, number 4,

Harvey, David, 2003, The New Imperialism (Oxford University).

Harvey, Mark, 2001, Undermining Construction: The Corrosive Effects of False SelfEmployment (Institute of Employment Rights).

Held, David, Anthony McGrew, David Goldblatt and Jonathan Perraton, 1999, Global Transformations (Polity).

Henderson, Jeffrey, 1989, The Globalisation of High Technology Production (Routledge).

Henwood, Doug, 1998, Wall Street (Verso),

Henwood, Doug, 2006, “The ‘Business Community’”, Socialist Register 2006: Telling the Truth.

Herod, Andrew, 1991, “Local Political Practice in Response to a Manufacturing Plant Closure: How Geography Complicates Class Analysis”, Antipode, volume 23, number 4.

Herod, Andrew, 2000, “Implications of Just in Time Production for Union Strategy: Lessons from the 1998 General Motors—United Auto Workers Dispute”, Annals of the Association of American Geographers, volume 90, number 3.

Hirst, Paul, and Grahame Thompson, 1999, Globalization in Question, second edition (Polity).

Holloway, John, 1994, “Global Capital and the Nation State”, Capital and Class 51 (summer 1994).

Holloway, John, 1995, “Capital Moves”, Capital and Class 57 (autumn 1995).

Hughes, Steve, 2002, “Coming in from the Cold: Labour, the ILO and the International Labour Standards Regime”, in Rorden Wilkinson and Steve Hughes (eds), Global Governance (Routledge).

Huws, Usula, 1999, “Material World: The Myth of the Weightless Economy”, Socialist Register 1999: Global Capitalism vs Democracy.

Hyman, Richard, 1992, “Trade Unions and the Disaggregation of the Working Class”, in Mario Regini (ed), The Future of Labour Movements (SAGE).

Hyman, Richard, 1999, “Imagined Solidarities: Can Trade Unions Resist Globalisation?”, in Peter Leisink (ed), Globalisation and Labour Relations (Edward Elgar).

ILO, 2005, “Promoting Fair Globalization in Textiles and Clothing in a Post-MFA Environment” (ILO Sectoral Activities Programme),

Kenwood, George, and Alan Loughheed, The Growth of the International Economy: 18201990 (Routledge).

Lash, Scott, and John Urry, 1987, The End of Organized Capitalism (Polity).

Lash, Scott, and John Urry, 1994, Economies of Signs and Space (SAGE).

Lenin, Vladimir, 1975 [1902], What is to be Done? (Foreign Languages Press),

Lewchuk, Wayne, and David Robertson, 1997, “Production without Empowerment, Work Reorganisation from the Perspective of Motor Vehicle Workers”, Capital and Class 63 (autumn 1997),

Piore, Michael, and Charles Sabel, 1984, The Second Industrial Divide: Possibilities for Prosperity (Basic Books).

Pollard, Jane S, 2005, “The Contradictions of Flexibility: Labour Control and Resistance in the Los Angeles Banking Industry”, Geoforum, volume 26, number 2.

Pollert, Anna, 1988, “The ‘Flexible Firm’: Fixation or Fact?”, Work, Employment and Society, volume 2, number 3.

Radice, Hugo, 1999, “Taking Globalization Seriously”, Socialist Register 1999: Global Capitalism vs Democracy.

Reich, Robert B, 1991, The Work of Nations (Simon & Schuster).

Rinehart, James, 1999, “The International Motor Vehicle Program’s Lean Production Benchmark: A Critique”, Monthly Review, January 1999.

Rinehart, James, Christopher Huxley and David Robertson, 1997, Just Another Car Factory: Lean Production and its Discontents (ILR).

Ross, George, 2000, “Labor Versus Globalization”, Annals of the American Academy of Political Science 570.

Rowley, Chris, and John Benson, 2000, “Global Labor? Issues and Themes”, Asia Pacific Business Review, volume 6, numbers 3 and 4.

Rowthorn, Robert, and Ramana Ramaswamy, 1997, “Deindustrialization—Its Causes and Implications”, International Monetary Fund,

Sabel, Charles, 1984, Work and Politics (Cambridge University).

Smith, Tony, 2000, Technology and Capital in the Age of Lean Production: _A Marxian Critique of the “New Economy” (State University of New York).

Strange, Susan, 1996, The Retreat of the State: Diffusion of Power in the World Economy (Cambridge University).

Sutcliffe, Bob, and Andrew Glyn, 1999, “Still Underwhelmed: Indicators of Globalization and Their Misinterpretation”, Review of Radical Political Economics, volume 31, number 1.

Thelen, Kathleen, and Ikuo Kume, 1999, “The Effects of Globalization on Labor Revisited: Lessons from Germany and Japan”, Politics and Society, volume 27, number 4.

Thieblot, Arman, 2002, “Technology and Labor Relations in the Construction Industry”, Journal of Labor Research, volume 23, number 4.

Thomas, Kenneth P, 1997, Capital beyond Borders: States and Firms in the Auto Industry, 196094 (Macmillan).

Thun, Eric, 2008, “The Globalization of Production”, in John Ravenhill, Global Political Economy (Oxford University).

Tilly, Charles, 1995, “Globalization Threatens Labor’s Rights”, International Labor and Working Class History, volume 47, number 1.

Walker, Richard A, 1999, “Putting Capital in its Place: Globalization and the Prospects for Labor”, Geoforum, volume 30, number 3.

Walton, Mary, 1997, Car: A Drama of the American Workplace (WW Norton & Company).

Waterman, Peter, 1999, “The New Social Unionism: A New Union Model for a New World Order”, in Peter Waterman, and Ronaldo Munck (eds), Labour Worldwide in the Era of Globalisation: Alternative Models in the New World Order (St Martin’s),

Weiss, Linda, 1999, The Myth of the Powerless State (Cornell University).

Wills, Jane, 2001, “Community Unionism and Trade Union Renewal in the UK: Moving Beyond the Fragments at Last?”, Transactions of the Institute of British Geographers, volume 26, number 4.

Wills, Jane, and Melanie Simms, 2004, “Building Reciprocal Community Unionism in the UK”, Capital and Class 82 (spring 2004), Analysis ,