The nature of de-industrialization

Initial accounts of de-industrialization in Britain saw it as a problem of the national economy, and discussion focused on the output and macro-economic, especially balance of payments, aspects.Footnote ¹ When a major public debate using this term erupted in the 1970s, the predominant view amongst economists at that time was summarized by Cairncross, who argued that de-industrialization ‘is a matter for concern if it jeopardises our eventual power to pay for the imports we need’.Footnote ²

This focus on the national economy meant little regard was had at that stage to the spatial aspects of the problem. As Martin and Rowthorn noted, it was the American de-industrialization literature of the early 1980s that brought this dimension to the fore, especially because of its concerns with ‘capital flight’ within the US economy, and the consequent flow of jobs from the north-east ‘rust-belt’, to the south and west ‘sun-belt’.Footnote ³

The concern with the balance of payments consequences of de-industrialization still animates some discussions, but it has become harder and harder to argue that the current account deficit, which has grown almost continuously since the 1980s, acts as a major constraint on policy as envisaged in the debates of the 1970s. Indeed, despite this burgeoning deficit, the balance of payments has largely disappeared as an economic policy concern over the last 30 years.Footnote ⁴ Rather, since the early 1980s, not least because of the employment collapse at the beginning of that decade, the British literature followed the American in placing much more emphasis on the employment consequences of de-industrialization, and on the spatial aspects of those consequences.

This article starts from the position that we should indeed focus on changing patterns of employment because of their links to economic welfare.Footnote ⁵ The post-war period, despite the hopes of some, has not seen the demise of the reliance of the great bulk of the population on wage income.Footnote ⁶ There has been a clear failure to move to ‘asset-based’ welfare, and the project of a ‘property-owning democracy’ espoused by the Right and some elements of the Centre-Left has turned out in practice to have results almost entirely in housing.Footnote ⁷ While home ownership has given economic security to some, it has distributed that security unevenly, and may have contributed to macro-economic instability.Footnote ⁸ So the traditional understanding that under capitalism most people continue to rely on waged work for economic welfare remains accurate.

What has been the effect of de-industrialization on the labour market? In summary, it has meant the radical decline in activities that offered large amounts of regular, relatively well-paid employment to working-class people with relatively limited educational qualifications (but often with high levels of job-specific skills). The now dominant service sector, in contrast, is characterized by a much more polarized range of jobs, with formal educational qualifications much more important in determining who gets the ‘lousy or lovely’ jobs.Footnote ⁹ This polarization is associated not just with great divergences in wage levels, but also less security for those at the bottom end; work for many has become much more uncertain and precarious.Footnote ¹⁰

In making this distinction between an ‘industrial’ and a ‘service’ dominated labour market we must be careful to historicize and not idealize the former. Historically, industrial labour has not always been characterized by stability, and it is paradoxical perhaps that the relative stability and prosperity of work in this sector is most evident in the post-war ‘Golden Age’, even as the sector's role in the economy passed its peak. So the positive features of industrial labour were in part contingent on wider circumstances. These included the macro-economic environment, but also the ‘virtuous circle’, whereby prosperous industrial work aided the effectiveness of trade unions, and effective trade unionism helped to underpin better conditions for workers in industry.

The standard categorizations of ‘industry’ and ‘services’ are not directly related to the labour market, so the fit between these categories and the employment characteristics we are interested in is by no means complete. For example, we should include in our definition of ‘industry’ employment in railways and utilities, which fall outside the most commonly used definition of industry. In addition, there are activities in the service sector which have some of the key characteristics of ‘industrial’ work (for example that done by maintenance workers in schools, hospitals and universities). Conversely, some kinds of less-skilled industrial work have always been casualized and precarious, and the incidence of this may have risen in line with the general weakening of demand for unskilled labour. Despite these qualifications, in seeking to understand broad changes in the labour market, ‘de-industrialization’ is the most useful characterization. Alongside its direct contribution to greater inequality and insecurity, it has contributed significantly to higher unemployment in the older industrial regions, much of this disguised by various kinds of disability and sickness benefits.Footnote ¹¹

De-industrialization has also had two effects on the public finances. On one side, higher unemployment in the older industrial regions has increased spending on disability and sickness benefits, even as the incomes of those reliant specifically on unemployment benefits (‘Job Seekers Allowance’) have been relentlessly squeezed since the 1980s.Footnote ¹²

But the counterpart to the successful attack on the incomes and standing of the unemployed has been the desire to ‘make work pay’ at a time when labour market polarization has pushed increasing numbers into levels of poverty, wages deemed unacceptably low. Hence the enormous growth in in-work benefits (alongside housing subsidies), creating a wage-subsidy system costing around £30 billion per annum. After 2010, and seeking to cut these subsidies in the name of ‘fiscal consolidation’, this ‘New Speenhamland’ system has brought a striking volte-face in Conservative politics, with the attempt to switch some of the cost of raising sub poverty level wages on to employers by a statutory ‘Living Wage’.Footnote ¹³ But this shift in policy, and some trimming of benefit levels, has only slightly dented the public expenditure cost.

So the case for focusing on employment change as the key problem of the national economy is a powerful one, and the general pattern is readily apparent in Table 1. But the spatial story, the distribution of sectoral employment change within Britain, is more complex. Much of the British literature of the 1980s tended to assimilate de-industrialization into an older story of regional disparities in economic prosperity, the ‘north–south divide’ between those areas still suffering the legacy of the decline of the nineteenth-century ‘staple’ industries, and those where the ‘new’ industries of the mid-twentieth century flourished.Footnote ¹⁴

Table 1.

Proportion of workers in industrial employment in the UK

Source: ONS databank.

But this incorporation of de-industrialization into a long-standing regional dichotomy appears problematic on two counts. First, the evidence is clear that de-industrialization began in the cities, cities distributed throughout the regions of Britain, not least in London in the ‘prosperous’ south-east.Footnote ¹⁵ Second, a large part of the regional disparities in rates of de-industrialization have been a function of how urban those regions were: ‘The decline in the number of manufacturing jobs in cities and the relative growth elsewhere has redistributed jobs between regions. Highly urban regions have declined; more rural regions have grown.’Footnote ¹⁶ Table 2 shows this urban/rural divide in the 1960s and 1970s, before the recession at the beginning of the 1980s in which manufacturing employment fell everywhere. It makes clear that trend in de-industrialization was hitting urban areas of all kinds, except small towns.Footnote ¹⁷ But the generality of the process, and the ‘surprising’ evidence of de-industrialization of cities generally never thought of as ‘industrial’, does not mean urban Britain had a homogeneous experience. To examine the spatial diversity of that experience, we look in detail at three very different urban cases.Footnote ¹⁸

Table 2.

Manufacturing employment change, 1960–1978, by type of area

Source: S. Fothergill, G. Gudgin, M. Kitson and S. Monk, ‘The de-industrialisation of the city’, in R. Martin and R. Rowthorn (eds.), The Geography of De-industrialisation (London, 1986), 231.

Three cases

Dundee is a classic ‘post-industrial’ city, once seen as an epitome of industrial Scotland, now dominated by services but with a continuing legacy of industrial decline. The population of the city, expanding slightly and then stable in the 1950s and 1960s at around 170,000, fell in every census after 1971 down to 2001 (with a total decrease of 20 per cent), before stabilizing in 2011. This smaller population exhibited many signs of the effects of de-industrialization, with an ageing population, high levels of unemployment and poor scores on the Index of Multiple Deprivation.Footnote ¹⁹

But the city's trajectory of de-industrialization has been far from straightforward, with effectively two separate ‘waves’ of that process. Before World War I, Dundee's economy was dominated to an extraordinary extent by one industry, jute manufacturing. Drawing on raw jute from Bengal, and selling its products all over the world, ‘Juteopolis’ was a striking example of the process of ‘imperial globalization’ which characterized much of the British economy in the decades before 1914.Footnote ²⁰ The industry was subject to increasingly serious competition from India's jute manufacturers from the 1880s, and by the 1930s the industry had shrunk to a fraction of its pre-1914 size. Protection from 1939 stabilized the industry, and thereafter until the late 1960s employment fell only slowly. Initially, the post-1945 shrinkage in labour demand from jute was more than offset by a successful policy of attracting new manufacturing companies, especially American multinationals. Regional subsidies to these companies were crucial in delivering an unemployment rate below the Scottish average in the 1950s and 1960s. However, the tide turned sharply in the 1970s. Protection for jute was dismantled, and the industry also faced increasing competition from the artificial fibre polypropylene, a much less labour-intensive product. Simultaneously, there began the ‘retreat of the multinationals’.Footnote ²¹ In Dundee, this was especially evident in the decline in employment at two large American companies producing electronic goods: Timex, where employment peaked at over 6,000 workers in 1974, declining to complete closure by 1993; and NCR, where employment peaked in 1970, and subsequently shrank to a residual, non-manufacturing, presence.Footnote ²²

Two spatial aspects of Dundee's development should be noted. First, the new jobs created there by the American multinationals were not based on ‘capital flight’. Though both Timex and NCR were based in the American north-east, the investment in Scotland was not a reallocation of resources away from the USA, but a net addition to activity aimed at capturing some of an expanding European market for the companies’ products. Similarly, the demise of these activities was not primarily a case of a shift to more profitable locations, but largely the result of technological changes that undermined the companies’ market position, with, for example, Timex's production of mechanical watches.Footnote ²³

De-industrialization is most dramatically evident in the process of closure of factories and mines. Understandably, the traumas of such closure, and the political fights that have been a common accompaniment, have dominated much of the literature on de-industrialization.Footnote ²⁴ But, of course, as a process, de-industrialization is about the limited creation of new industrial jobs as much as the loss of the old. When industrial employment flourished, some jobs were constantly being lost, but there were plentiful replacements. But with de-industrialization, ‘deaths’ outnumbered ‘births’ and the population of jobs declined. This pattern has been clearly identified in Glasgow, where the decline of industrial employment in the city was, quantitatively, as much the result of the lack of new job creation as the loss of the old.Footnote ²⁵

More broadly, of course, the impact of de-industrialization depends upon what jobs, if any, fill the gap. In Dundee, a large part of the answer was the public sector, with the National Health Service and higher education, along with the local authority, emerging as the biggest employers. This was a common pattern in the ‘post-industrial’ regions. In other parts of the country, the public sector has played a smaller role, and private sector services a much larger one.Footnote ²⁶ This is perhaps most evident in the case of London.

London might be seen as a long way from Dundee in every regard. But both historically were cities of industry. In 1961, Peter Hall emphasized that London had been the most important manufacturing centre in Britain for as long as records exist. This position rested upon significant shifts in industrial composition, from the historically long-standing importance of industries such as clothing, furniture and printing, to the growth of engineering and vehicles strongly evident in the inter-war years.Footnote ²⁷ But even before the publication of Hall's landmark book, the number of industrial jobs in London was in decline. Manufacturing employment in London peaked in the 1950s, and from the 1960s this trajectory was greatly added to by the loss of jobs in the docks, and many of the industries associated with the port of London.

By the early 1980s, the Greater London Council's London Industrial Strategy was predicated upon an analysis which saw London as suffering an acute process of de-industrialization, without a compensating rise in jobs in other sectors, resulting, they estimated, in over half a million Londoners without jobs: ‘the largest urban concentration of unemployed people in the advanced industrial world’.Footnote ²⁸ The Strategy also noted some of the changing localized geographical impacts of the process. In the second half of the 1960s and through the 1970s, it was concentrated in inner London and the east, linked especially to the accelerating decline of docklands in the face of containerization.Footnote ²⁹

De-industrialization was thus integral to the ‘inner-city’ problem which became central to discussions of urban Britain from the late 1960s.Footnote ³⁰ Part of the population decline of London and other major conurbations in the early post-war decades was the result of deliberate policy decisions about slum clearance and dispersal, aimed at shifting both populations and employment to outer suburbs or new towns.Footnote ³¹ The presumption of these policies was most of this employment would be in industry, especially ‘light manufacturing’, seen as much less location critical than the ‘old staples’. The adverse impact of these policies for the centre of cities was recognized from the late seventies, though the debate about the scale and nature of these effects continues.Footnote ³²

In the early 1980s, whilst the problems of the inner-city areas were exacerbated, serious de-industrialization took place in west London, especially concentrated in the ‘new’ engineering activities which dominated the industrial estates of this part of London.Footnote ³³ In 1981, 19.2 per cent of all jobs in London were in manufacturing, but in the following decade, this fell to 11 per cent, the absolute figure falling from almost 700,000 to 359,000. Over the same period, total employment in London fell by 300,000.Footnote ³⁴ However, manufacturing output in London fell only slightly in this period, emphasizing the divergence between employment and output trends when productivity is growing fast. Also with broader significance is the fact that over the same period while productivity in banking, finance and business services grew rapidly (along with output and employment), in other service sectors productivity grew much more slowly (though faster than in comparable sectors elsewhere in Britain). This set a pattern for London which was maintained up to the 2008 crisis: sharply rising productivity in manufacturing accompanying a rapid fall in employment; a rapid rise in employment, output and productivity in finance; rising output and employment, but slow productivity growth, in services.Footnote ³⁵

By 1991, London had a lower proportion of its workforce in manufacturing than the national average. Viewed positively, this can be seen as a result of the buoyancy of the financial services sector, though there was also expansion in the lower productivity (and lower-waged) parts of service activity. But this shrinkage went along with a falling total working population and the evidence of considerable economic problems in parts of the city.Footnote ³⁶

Some critics of the London Industrial Strategy disputed the view that London suffered from a general process of de-industrialization, arguing that the capital's loss of such employment was the result of relocation of its industry to elsewhere in Britain, rather than part of a general de-industrialization of the whole country.Footnote ³⁷ In part, this argument is undoubtedly true. As already noted, London's share of employment in manufacturing fell below that of the UK as a whole by 1991, so its experience was divergent from the rest of the country. And relocation rather than simple closure was an important factor in London's industrial experience. Even in the early post-war years, when total employment in manufacturing was expanding, shifts of activity to locations outside London were commonplace. A study of north-west London found that in that area between 1940 and 1964, 147 factories closed down completely, moving all activity outside the city.Footnote ³⁸

The forces at work in such relocations seem clear. First, were the physical limits to expansion in crowded urban areas, followed by higher rents. For example, rents per square foot on the huge Park Royal industrial estate in Acton and Willesden rose tenfold between 1939 and 1964. Second was the very tight labour market in the early post-war years, resulting in upward pressure on wage costs and high labour turnover. Finally, there were the effects of government policy. While regional policy after 1945 greatly encouraged the expansion of industrial employment in Dundee (and much of the rest of ‘industrial Britain’), post-war urban planning actively discouraged expansion of such activities in London, by refusing the necessary building permissions.Footnote ³⁹

Relative to a post-industrial city such as Dundee, London has clearly shown greater capacity to deal with the effects of de-industrialization. Private sector service activity has been much more buoyant, and the role of the public sector as an employer has been much lower. But inequality and poverty remain major features of the city, even as the geography of deprivation has become less concentrated in the inner city.Footnote ⁴⁰

The relocation of industry from London bolstered manufacturing employment elsewhere in Britain during the early post-war years. One beneficiary was the town of High Wycombe, in South Buckinghamshire, about 30 miles outside London. In 1951, Rowntree and Lavers were clear: ‘High Wycombe is an industrial town; by far the most important industries are furniture and chair-making.’Footnote ⁴¹ The role of furniture in Wycombe was not new, the industry tracing its roots back to eighteenth-century chair-making, using timber from the local beech woods.Footnote ⁴² Employment in the industry reached its peak in 1955, at just under 10,000 workers, around 25 per cent of the town's workforce.Footnote ⁴³ This employment was bolstered by firms migrating from (central) London from as early as the late nineteenth century. But migration was probably at its peak in the early post-war years; London's share of employment in the industry was still predominant (over 40 per cent) but was falling in the 1950s, whilst that of High Wycombe rose.Footnote ⁴⁴ Typical was the shift of factories from Tottenham, a traditional centre of the furniture industry in north London.

Although a very different town in many respects, the industrial trajectory of High Wycombe after 1945 had something in common with Dundee. In Dundee, the staple industry, jute, declined (slowly until the 1970s) but there was expansion in engineering; in Wycombe the ‘staple’ of furniture also peaked in the 1950s and declined only slowly over the next two decades. Engineering employment grew in both places, though this began earlier than in Dundee, and owed much to wartime attempts to escape the London blitz, rather than to post-war regional planning. From 1938 to 1958, employees in engineering in Wycombe expanded in number from 900 to 6,500, and continued to increase thereafter, reaching a peak of around 16,000 in the late 1970s.Footnote ⁴⁵

As with furniture, some of this growth in engineering came as a result of relocation from London, both from the desire to escape wartime bombing, and after 1945 from the cost concerns and physical controls arising from regional policy. Whilst located just outside the town's boundaries, the shift of the precision-engineering tobacco-machinery company, Molins, from Deptford in south-east London to Wycombe in 1950, was typical of the process.Footnote ⁴⁶

Like all of industrial Britain, Wycombe was hit extremely hard by the exchange rate appreciation of 1979–81, which greatly encouraged the acceleration of the trend for the trade balance in furniture to decline, the industry moving into a large trade deficit from1981. (Until the 2000s, the great bulk of these imports came not from low-wage Asian countries, but from Italy, Belgium, the Netherlands, Germany and Czechoslovakia.)Footnote ⁴⁷ Engineering employment was also hit hard. But as elsewhere in Britain, this sudden decline only accelerated a long-term trend towards a shrinking of industrial employment. By the end of the twentieth century, High Wycombe shared some of the prosperity of the commuting towns surrounding London, and a low unemployment rate, combined, like Dundee, with an ageing population and continued shrinking of industrial employment.Footnote ⁴⁸

These three cases emphasize that the post-1945 history of very different types of urban areas have all been seriously affected by de-industrialization. The precise patterns, forces at work and chronology have differed, as have the long-run impacts on economic welfare. But no part of Britain has been immune from its effects.

Narratives of post-war Britain

De-industrialization is one of several narratives which have been widely deployed to analyse post-war Britain. Three others are of particular importance: economic decline; the ‘triumph’ of neo-liberalism; and, for the most recent decades, the rise of globalization. The contention here is not, of course, that these three alternative frameworks are valueless, but rather that they are best seen as less fruitful for our understanding than de-industrialization.

From when it was (re)invented in the late 1950s until the 1980s, economic declinism was central to contemporary politics in Britain, finding support across the political spectrum. This political positioning in turn stimulated a variety of narratives that purported to show the historical roots of this decline, similar to the way in which earlier political debate had stimulated declinist histories.Footnote ⁴⁹ Since the 1980s, declinism has been less evident in political debate, but still survives in journalistic and academic writing.Footnote ⁵⁰ Nowadays, the narrative is mainly supported from the political Left; those on the Right, and indeed in the Centre, tend to adhere to the view that Thatcherism ‘reversed decline’ and therefore no longer see it as a key problem.Footnote ⁵¹ Historical declinism still has its adherents, but the narrative does not dominate as it once did.Footnote ⁵²

The complex roots and the weaknesses of declinism as a historical narrative have been explored elsewhere, and these arguments will not be repeated here.Footnote ⁵³ Suffice to say that, if we restrict ourselves to the highly conventional (but deeply problematic)Footnote ⁵⁴ use of GDP growth as the key measure of economic performance, ‘decline’ seems an odd word to use to describe long-run British economic performance. According to IMF data for 2015, Britain is twenty-fifth in the world ranking of GDP/head, the only large countries with a higher level being the USA (eleventh) and Germany (eighteenth).Footnote ⁵⁵ The USA overtook Britain in the late nineteenth century, so the only large country to overtake Britain post-1945 is Germany, where in 2015 GDP/head was around 12 per cent higher. For a historian, the German comparison is especially resonant. In the book Made in Germany in 1896, the impending doom of Britain was spelt out; 120 years later the difference in GDP/head barely exceeds the measurement error.Footnote ⁵⁶ On a global scale, Britain worrying about ‘decline’ is akin to a millionaire complaining that he is not a multi-millionaire; the great bulk of the world's population have less than a quarter of Britain's GDP/head.

If we are focused on the national economy, the point is not, of course, that Britain has not and does not face significant economic problems (like almost all economies), but that the ‘decline’ literature offers limited insights into the nature and causes of these problems, not least because of its tendency to seek for moralistic and ideologically charged stories about who is to blame for economic adversity. When we turn to the spatial differentiation discussed in our three examples above, the notion of ‘decline’ as a pathological state seems even less helpful. In the case of Dundee, for example, while there have been occasional politically inspired attempts to find culprits for the difficulties of the local economy in the 1970s and 1980s in trade union intransigence, such as at the time of the prospective building of a Ford car plant in the city, investigation has shown that the great bulk of employers around that time found the city's labour force efficient and amenable.Footnote ⁵⁷

Neo-liberalism

‘Neo-liberalism’ is a term rooted in economics.Footnote ⁵⁸ Most accounts see it as combining elements of a number of strands of thought, most importantly neo-classical economics, ordo-liberalism and Austrian economics.Footnote ⁵⁹ The addition of public choice theory added an extra, powerful, component. While the original historiography of this ideological movement emphasized the role of think-tanks like the Institute of Economic Affairs, we now know that the spread of the doctrine was aided from the early post-war years by its allies in business and politics.Footnote ⁶⁰ But it was only in the crisis years of the 1970s that it gained much influence. That influence cannot be understood except in the context of the circumstances of that decade, and how the experience of rapid inflation and ballooning public deficits was constructed as an indictment of ‘the Keynesian welfare state’. In these circumstances, a reinvigorated, right-wing, declinist narrative opened the space for the new pro-market, neo-liberal, doctrines to flourish.Footnote ⁶¹

But how far did this doctrinal flourishing turn into policy success? While undoubtedly policy did change from the end of the 1970s, how far was that change deep enough to justify the view that neo-liberalism ‘triumphed’? At its most general, neo-liberalism is characterized by a belief in the moral, political and economic superiority of market mechanisms over the state and politics in organizing society. While (some) neo-liberals may wish to strengthen the state in specific areas (anti-monopoly laws, policing, the military), central to the political aims of neo-liberalism has been an overall ‘rolling-back’ of the state. One key measure of that ‘rolling-back’ are trends in public spending. These show that, whatever their intentions, governments have struggled to cut that expenditure. As Martin Daunton stresses, ‘Public expenditure was 42.2% of GDP when Mrs Thatcher came to power, and 42.3% in 1995/6. Mrs Thatcher did not roll back the state.’Footnote ⁶² In the Thatcher/Major years, absolute levels of expenditure grew significantly. Under New Labour, despite initially following highly restrictive policies, in the new millennium public expenditure shot up. As a result, by the eve of the financial crisis in 2007, public spending had grown to 240 per cent of its level in 1970; there had been no fiscal ‘race to the bottom’ (see Figure 1).

Figure 1.

Public expenditure 1970/71–2007/08 (total managed expenditure in £b at 2008/09 prices)

Source: ONS databank.

Perhaps even more significantly, not only has total spending risen, but the proportion of this total spent on the welfare state has grown. Spending on health care and education have both risen as a share of GDP. In social security, while the aim of squeezing the incomes of the unemployed has been carried through, overall social security spending has risen in line with GDP.Footnote ⁶³ This fact is particularly significant when one notes the centrality of anti-welfare state ideas to all types of neo-liberalism. The precise desired limits to state welfare provision were disputed amongst neo-liberals in the early years of the development of the doctrine, but the kind of expansion seen from the 1960s was unanimously anathemized.Footnote ⁶⁴ In Britain, the neo-liberal think-tank the Institute of Economic Affairs ‘has stood out for the strength of its attack on the so-called “welfare state”’, and this issue dominated its early publications.Footnote ⁶⁵ Yet in a spending sense, we have since the 1970s moved from a ‘warfare state’ to a truly ‘welfare state’, a paradoxical consequence if one believes neo-liberalism has been triumphant.Footnote ⁶⁶

Equally, for all the talk in the 1980s of the defeat of Keynesianism by monetarism, fiscal policy, while strongly repudiated in the crisis of the early 1980s, was not permanently defeated. As the initial expansionary fiscal responses to the crises of the early 1990s and 2008–10 demonstrate, fiscal stimulation in time of recession remained alive and well.Footnote ⁶⁷ Commenting on the Thatcher period, Peter Taylor-Gooby remarked that: ‘One verdict on Thatcher's legacy might be that the policies failed but that the ideology grows ever stronger.’Footnote ⁶⁸ This suggests a wider problem in discussing neo-liberalism in Britain; its undoubted ideological virulence cannot be taken as a straightforward measure of its implementation.Footnote ⁶⁹ This is not to dispute that neo-liberal ideas have been significant ideological underpinnings to some hugely important changes in post-1970s Britain. Labour market de-regulation coupled with anti-trade union policies have helped to engineer major changes in the economic and political landscape. Alongside this, as Avner Offer has persuasively argued, one of the most striking and significant economic (and social) changes in Britain since the 1970s has been the multiplication of private debt (which now finances not only house purchases but the great bulk of new cars as well as other consumer items). He argues that this is linked to a switch away from a broadly social democratic notion of collective risk-bearing, to one of individuals having to provide as best they can for life's contingencies, which has commonly meant household debt accumulation, in turn an important aspect of the wider phenomenon of ‘financialization’, a key shift in corporate structure and behaviour.Footnote ⁷⁰

Alongside the multiplication of debt, the outstanding effect of the change in policy at the end of the 1970s was the profound increase in inequality in the following decade.Footnote ⁷¹ The scale of this shift seems to have resulted from a conjunction of forces: as Hills notes, ‘most of the factors determining income distribution were pushing in the same direction’.Footnote ⁷² But while the shift was not all policy-driven, there is no doubt that neo-liberal policies made a substantial contribution, and this was consonant with the ideological assault on the pursuit of greater equality by social democracy.

The key problem with the ‘triumph of neo-liberalism’ as a general narrative about post-war Britain is that its success as a policy programme has been seriously constrained by changes in circumstances, especially changes in industrial structure and the labour market, almost entirely beyond its control. As argued in the first section, de-industrialization in part derives its explanatory strength from the understanding it can offer about how economic change has driven supposedly ‘neo-liberal’ governments both to increase subsidies enormously in the labour and housing markets, and to increase state intervention greatly in that market by a statutory, national minimum wage. We can summarize this by saying that de-industrialization has acted as a major constraint on the realization of any neo-liberal programme of ‘rolling-back the state’.

Neo-liberalism is too broad a term to give us much insight into the spatial differentiation of Britain's urban economies. In an ultra-centralized polity like modern Britain, urban areas have found it extremely difficult to pursue distinctive, non neo-liberal economic programmes. The important analysis of London's problems in the London Industrial Strategy was noted above, but the body that threatened to follow up such analysis with strategies of public intervention, the Greater London Council, was abolished by central government before those strategies had time to develop.

Globalization

While declinism and ‘triumph of neo-liberalism’ narratives have mainly been deployed to analyse the national economy, globalization has been used to seek to explain geographical differences. The Brexit decision in June 2016 stimulated an argument that the vote could be explained in large part by economic factors, the creation of a significant section of the population who had been disadvantaged by the spatially uneven economic consequences of globalization. Proponents of this view ranged from journalists to Christine Lagarde, managing-director of the IMF.Footnote ⁷³ One obvious qualification to such claims, linking economic disadvantage to Leave voting, is that the (small) Leave majority was most strikingly characterized by its greater age and weaker educational profile, rather than directly by economic condition.Footnote ⁷⁴ The decision to leave the EU was a clear victory for traditional Conservative voters, many of whom live in the prosperous south of England, rather than in less-well-off parts of the country; two-thirds of those who voted Labour in 2015 voted Remain. Overall, we can perhaps helpfully characterize the Leave vote as an alliance between those who favoured ‘taking back control’, and those who saw the vote as a means of expressing opposition to their economic difficulties.

But insofar as Brexit reflected underlying economic discontent in some parts of the UK, is it helpful to link this to the effects of globalization? Globalization is multidimensional, involving freer movement of goods, capital and (to a much lesser extent) people, and has complex implications. But in terms of its most important economic effects on the British population, the argument made here is that in seeking to understand the economic basis of the Brexit vote, we should concentrate not on globalization but on the long-term impact of de-industrialization. It has been that process of de-industrialization, outlined in the first section, that has underpinned much of the growth of sharp spatial divisions in economic conditions across Britain.

The post-Brexit debate tended to conflate these two issues. For example, Diana Coyle wrote ‘The UK's “Leave” vote could be seen as a vote against globalization and its uneven impact on different parts of the country, rather than a vote specifically against the EU. The proportions voting for Leave were higher in the Midlands and North of England, where de-industrialization struck hardest and where average incomes have stagnated.’Footnote ⁷⁵ Such claims reflect a common error of conflating globalization and de-industrialization, a confusion which ignores the clear historical evidence of the way de-industrialization began long before the current phase of globalization.

Globalization has contributed to de-industrialization, but it is only one contributor, and historically not the most important. As shown in Table 1, de-industrialization began in Britain in the 1950s. It was driven by shifts in patterns of demand and technological change, most strikingly in increasing the growth of productivity (and lowering the relative price) of manufactured goods. Since 1948, the annual increase in labour productivity in services has been around 1.48 per cent, in manufacturing 2.78 per cent. (‘Industry’ is normally defined in the official statistics as manufacturing, plus mining plus construction. Mining has almost disappeared, but construction has come to employ almost as many workers as manufacturing; productivity growth in construction is significantly slower than in manufacturing.)

Note that in that same period, industrial output has not fallen, but grown slowly on trend, so it is now at an all-time high. These employment trends have affected all industrial countries, so that industrial employment has fallen substantially even in successful industrial countries with a manufacturing trade surplus, such as Germany.Footnote ⁷⁶ Industrial employment there has fallen more slowly than in Britain, but as a share of the total has more than halved since its peak in 1970.Footnote ⁷⁷

The sustained nature of these changes, which long pre-dated the late twentieth-century surge in globalization, can be illustrated with some examples. Take the pattern for coal, shown in Figure 2. The longevity of these trends is illustrated by the fact that many more coal-mining jobs were lost in Britain under Harold Wilson's government of the 1960s than under Margaret Thatcher's government of the 1980s. These trends reflected the collapse in industrial and domestic uses of coal, and latterly the shift away from coal in electricity generation.

Figure 2.

Employment in coal in the UK, 1921–2010 (000s of workers)

Source: M. Kerai, ‘Coal statistics in 2012’, www.gov.uk/government/uploads/system/uploads/attachment_data/file/170708/et_article_coal_in_2012.pdf (accessed 28 Apr. 2017).

Coal is important for our understanding of de-industrialization in a number of ways. The trajectory sketched above emphasizes the long-run nature of its decline, but also how much of this occurred during the ‘Golden Age’ of the 1950s and especially the 1960s. In this period, the impact of its decline was much mitigated by the ‘moral economy’ of the period, which meant that policy was predicated on the need to provide alternative jobs elsewhere in the coal industry, or alternative jobs in other sectors in the local economy.Footnote ⁷⁸ While the economic welfare effects of these shifts are complex, it seems clear that they were much less severe than in the years from the 1980s, when that moral economy broke down, and redundancy in the coalfields was often a route into unemployment or low-paid work, with coal-miners’ wives rarely able to ‘compensate’ household incomes by finding well-paid local employment.

It should also be noted that the moral economy of coal in the years down to the 1980s involved a nationalized industry acting, up to a point, as a ‘humane’ manager of decline. This role was hugely important in the nationalized industries in the 1950s and 1960s, because they saw some of the largest employment contractions. One calculation suggests that 80 per cent of the decline in employment in mining and manufacturing in north-east England between 1951 and 1981 was in the public sector, plus another 20,000 in nationalized railways.Footnote ⁷⁹ It is true that with the decline of the British coal industry (deep-mining of coal ended in 2016), Britain has become a coal importer, with 25 million tons of imports in 2015. But if we ask as simple counter-factual how many jobs would be created if that coal were produced at home, the number would be perhaps 12,500. This is not a trivial number, but it would only take employment back to the level of the year 2000, and, of course, leave the numbers a tiny fraction of the industry's previous size (and the main source of demand, coal-generation of electricity, is due to be phased out by 2025).

The history of the steel industry is also highly instructive. Figure 3 shows employment in this industry alongside output. Iron and steel has lost 95 per cent of its workforce since 1971, the great bulk of the fall coming in the 1980s (50 per cent of the fall in 1979–81). Over the same period of forty-five years, output fell much less, by approximately 50 per cent. Two points about this history should be stressed. First, Britain still produces a considerable amount of steel, but huge increases in productivity (and changes in the composition of output) mean this output can be produced with far fewer workers. Second, the fall in output owes little to the Chinese competition which has been so much emphasized in recent discussions. There was no big surge of Chinese imports until the 2000s; indeed, China was a steel importer well into the 1990s. Similar stories of long-run decline could be told of textiles and clothing and shipbuilding. But also important is the story of more characteristically ‘modern’ industries such as car manufacturing. Like iron and steel, output has diminished since the early 1970s peak, but has revived since the financial crisis to well over 90 per cent of that peak (though the composition of output has of course altered significantly). However, again because of productivity gains, far fewer workers are needed to produce a car than was the case 40 years ago. While today car manufacturing directly employs around 150,000 workers, in 1972, the figure was half a million for only slightly more output.Footnote ⁸⁰

Figure 3.

Output and employment in the UK steel industry, 1971–2015

Source: C. Rhodes, ‘UK steel industry: statistics and policy’, House of Commons Briefing Paper 07317, 28 Oct. 2016, 6, 9.

The key point here is that, if by some extraordinary circumstance, globalization, defined in relation to trade flows, was somehow reversed, it would not return Britain to having anything like the number of industrial jobs that existed in the 1950s (or 1980s). Some future increase in industrial jobs is of course possible, including by the expansion of such sectors as renewable energy. But the idea of a wholesale ‘re-industrialization’ which would offset the effects of the profound changes in demand and productivity (as well as globalization) of recent decades is fantastical.Footnote ⁸¹

Our three case-studies are also instructive on the limits of the globalization narrative in explaining their economic fortunes.Footnote ⁸² London, as noted, was a pioneer of de-industrialization, this being strongly evident from the 1950s, and long before the post-war revival of globalization had got into its stride. Insofar as this process involved the flight of jobs to other locations, those locations were overwhelmingly elsewhere in Britain (especially southern England), not to other parts of the globe. Globalization did play a part in the economic development in High Wycombe and Dundee, especially in the latter case and the demise of jute employment. But their overall economic development was more directly to do with technological change and rising manufacturing productivity than global integration per se.