The political economy of the service transition
The rise of the service economy creates new choices over policy and opens up opportunities for new coalitions adapted to the dynamics of post-industrial labour markets.
Over the past thirty years the wealthiest OECD economies – in Europe, North America and Australasia – have experienced rapid de-industrialisation. A range of factors have contributed to the de-industrialisation process: some, like technological change and changes in the characteristics of consumer demand, are internal to the development process in the economies themselves; others, like increased competition from developing countries in the market for manufactured goods, are external. Whatever its roots, there is no doubt that the impact of de-industrialisation on labour markets has been profound: more than three quarters of employment in most OECD countries is now in services, while industrial sectors, on average, account for less than one fifth.
This sectoral shift in the locus of economic activity has potentially radical implications for politics and society. In a new book on The Political Economy of the Service Transition (Wren, 2013; hereafter: PET), I have brought together a group of scholars from Europe and North America to assess the implications of the service transition for existing socio-economic regimes. We investigate how variations in the underlying institutional structures of alternative ‘varieties of capitalism’ have influenced their ability to manage the transition to services – in particular their capacity for creating new types of jobs in the face of declining opportunities in core manufacturing sectors, and the distributional outcomes with which new strategies for employment creation are associated. We also analyse the implications of the transition for politics – and for the sustainability of existing socio-economic models. The central findings of this volume speak to the on-going debate on the pages of Renewal and elsewhere about the viability of alternative ‘varieties of capitalism’ in a post-crisis world.
In the book we identify a set of critical challenges which de-industrialisation and the transition to services create in the economic and electoral arenas. In this essay, I will outline these challenges in brief, describing what I see as their primary implications for alternative models of capitalism.
The transition to services: economic challenges
One of the most serious socio-economic challenges posed by the de-industrialisation process is a reduction in job opportunities at medium and lower skill levels. The question for governments is: as the core of employment in manufacturing sectors shrinks, how, and where, are new jobs to be created? (Wren, chapter 1, PET). At stake in the resolution of this problem, also, is the future of the welfare state: reductions in the number of industrial workers in stable jobs who are contributors to the system threaten the affordability of welfare state provision, unless new jobs can be created in services (Nelson and Stephens, chapter 4, PET).
Low-skilled employment and the employment-equality trade-off
One ‘solution’ to the problem of declining employment opportunities at medium to low skill levels, which has been pursued most aggressively in Liberal regimes like the US and UK over the past few decades, has been to facilitate the expansion of low-skilled (and low-paid) private sector employment in personal, consumer, and social services. These countries have enjoyed considerable success in terms of employment creation in these sectors in recent decades. However, there are marked downsides to the strategy which they have pursued.
The first is that it has been heavily reliant on keeping relative wages in low-skilled service sectors low. The demand for personal and consumer services is very responsive to changes in prices (this is unsurprising when we consider their capacity for home production – think of catering and gardening services, for example). However, the capacity for productivity increases in the provision of these kinds of services is low (waitressing and childcare are good examples to think through here – it might be possible to increase the number of children supervised by one carer, the number of tables served by one waiter and so on, but in the process the quality of care and service will almost certainly decline). Given a low capacity for productivity growth, it becomes particularly important to keep relative wages in these sectors low, in order to generate a demand expansion based on the high price elasticity of demand for these kinds of services.
As a result it is harder to combine the expansion of lower-skilled service employment with equality than it was during the so-called ‘golden age’ of manufacturing expansion in the 1950s and 1960s (Iversen and Wren, 1998). (When the simultaneous occurrence of high demand elasticities for new consumer durables, and high rates of productivity growth in manufacturing sectors engendered by Fordist innovations in production processes, meant that relative prices could be kept low at the same time as real wage rates in these sectors were growing: Meidner, 1974; Rehn, 1985).
In Liberal regimes, then, the dominant response to the trade-off between wage equality and employment creation in low-skilled, low productivity service sectors has been to emphasise the goal of employment creation, with increases in wage inequality facilitated by the removal or reduction of protections on the wages of low-paid workers, attacks on the power of trades unions, and de-centralised wage bargaining.
As the experience of recent years has made clear, however, increasing inequality has not been the only downside of the Liberal strategy for employment growth over the past quarter century. In addition to these supply-side adjustments, there was a strong demand side component to the expansion of employment in low-skilled service sectors in these regimes during this period.
Part of this demand resulted from increases in family working hours (as more women entered paid work, and more workers began to work longer hours). Increases in working hours raise the demand for consumer and personal services via a substitution effect (women working in the paid labour force purchasing childcare and catered food, for example), but also via an income effect (families that work more hours earn more money, and personal and consumer services are ‘luxury’ items which occupy proportionately more of individual and household budgets as incomes rise) (Gregory, Salvadera, and Shettkat, 2007). Both of these effects can be seen as outcomes of structural labour market change, therefore, which are not inherently unsustainable.
By the turn of the century, however, the expansion of demand, and of employment, in these sectors was also closely linked to the expansion of cheap consumer credit and the asset (and in particular house) price inflation which accompanied it. Just how important this highly unsustainable element of the employment growth model had become in Liberal regimes by the early part of this century has yet to be accurately assessed: what is clear, however, is that the shock to employment in these countries in the wake of the crisis (when the supply of credit dried up) was relatively severe.
In contrast to the Liberal regimes, in the co-ordinated regimes of central and Northern Europe, levels of coordination in wage setting (and, as a result, levels of wage equality) remain relatively high, and the extent of reliance on employment creation in low paid, low productivity private service sectors in recent decades has been considerably lower. The social democratic regimes of Scandinavia, for example, have (to varying degrees) managed to continue to pursue simultaneously high rates of employment and wage equality by employing large numbers of workers at all skill levels in public service sectors. Meanwhile some continental European regimes – like Germany – have kept unemployment levels down in part by their continued strong performance in traditional export-oriented manufacturing sectors, but also because large proportions of the working age population (and large numbers of women in particular) remain outside the paid labour force.
The lower levels of reliance on debt-financed consumer demand in these regimes before the crisis meant that the shock which they experienced when the crisis hit (and credit dried up) was less severe. Meanwhile, the implications for equality of the pursuit of either of these strategies are also more positive than in the case of Liberal regimes – although the inegalitarian nature of strategies which rely on discouraging labour force participation amongst certain segments of the population should not be overlooked.
Critically though, none of these strategies is, on its own, a sufficient long-term solution to the problem of de-industrialisation. Without a thriving set of high value added sectors to finance them, the expansive public service sectors of Scandinavia are ultimately unaffordable. Without the existence of a core group of well-paid workers in high productivity export sectors, the large numbers of early retirees and women working within the home in Germany and other continental European countries cannot be supported. The strategy of expanding employment in low productivity private services, meanwhile, relies not just on low relative prices, but also on rising incomes and, as the Liberal experience of the past decade has shown, an over-reliance on the expansion of credit and ‘wealth illusion’, rather than productivity and income growth as a basis for the expansion of demand and employment in these sectors, is both unsustainable and economically costly.
Any sustainable strategy for employment growth must depend instead on the expansion of output and employment in high productivity sectors and, in the context of de-industrialisation, this means that expansion in high productivity service sectors is increasingly key. So what does this mean in terms of policy?
Skills and high-end service expansion
The revolution in information and communications technologies (ICT) which has taken place over the past three decades has radically transformed production and trade in certain areas of services. In so-called ‘knowledge-intensive’ sectors – business, finance, and communications, for example – the capacity for productivity growth and for trade have been hugely enhanced by rapid access to large amounts of information, and the ability to communicate that information – both locally and globally – rapidly and cheaply.
The transformation has not occurred in all areas of service production. In some sectors (for example the personal and social services discussed earlier) the impact of the new technology is less marked. Primarily this is because the provision of these services is, to varying degrees, reliant on face-to-face human interaction: a computer, for example, cannot increase the number of children that can be effectively supervised by a child care worker, nor can it facilitate her in providing this service to children who are not in her immediate vicinity (i.e. in participating in international trade). In these areas, then, the uptake of the new technology has been relatively low, and these remain essentially low productivity, non-traded sectors. In contrast, in the more knowledge-intensive sectors, where provision is less heavily reliant on face-to-face human interaction, and information and its transmission are at a premium, rates of productivity growth and international trade have increased rapidly.
Given their high capacity for productivity growth and for trade, these ‘dynamic’ service sectors have a critical role to play in any sustainable strategy for employment creation in the context of de-industrialisation. The important question, then, is what kinds of policies promote expansion in these types of sectors? In our book we place a heavy emphasis on skill formation, and on the capacity of different ‘varieties of capitalism’ to provide the skills which high-end service markets require. And here, the fact that these sectors are highly ICT-intensive has important implications.
It is by now well-established empirically that ICT and college educated labour are complements in production. As Autor, Levy, and Murnane (2003) point out, the new technology is highly effective at performing routine tasks which can be specified by stored instructions – even where the required programs are highly complex (for example, book-keeping or clerical work). As a result it acts as a substitute for labour in performing these tasks, which are typically carried out by workers at medium-skill levels (those with secondary, or some (but not complete) college education). It is less effective, however, at performing non-routine cognitive tasks requiring ‘flexibility, creativity, generalised problem solving, and complex communications’ (Autor, Levy, and Murnane, 2003, 5). Rather it serves to complement the skills of the (typically college educated) workers who perform those tasks: faster access to more complete market information, for example, may improve managerial decision-making, but it cannot substitute for that decision-making. Since technology is a complement to, rather than a substitute for, this type of human capital, then, investment in the new technology increases the demand for college educated labour.
Successful expansion in ICT-intensive service sectors is therefore reliant on the existence of an adequate supply of workers with tertiary-level skills. In this regard also the characteristics of the current era of service expansion differ significantly from those of the era of industrial expansion which preceded it. In the 1950s and 1960s, Fordist industrial expansion was associated with an increased demand for labour at low to medium skill levels – and was particularly notable for the existence of complementarities in production between low and high-skilled industrial labour (see, for example, Wallerstein, 1990). In contrast, successful expansion in high-end service sectors requires up-skilling, and increasing the numbers of workers receiving high quality tertiary education. This underscores, of course, the importance of ensuring effective investment at the tertiary level, and also in facilitating tertiary enrolment and access. Recent research indicates, however, that it also implies a critical role for investment in school based learning beginning as early as the pre-primary level, since education at this level is increasingly regarded as a key determinant of tertiary outcomes – especially for children from lower-skilled households (see, for example, Cuhne and Heckman, 2007; Heckman and Jacobs, 2010).
So how well equipped are existing ‘varieties of capitalism’ to meet the skills demands of the service economy?
Liberal regimes have been relatively successful thus far at producing large numbers of high quality college graduates. Levels of public educational investment are low in these countries in relative terms, however, and the system relies on the existence of high levels of wage inequality, which create strong incentives for private educational investment (since the relative rewards for such investment are high), and the extensive use of student loans.
One clear downside of the model for those concerned with equity of outcomes, therefore, is that it comes with high levels of wage inequality attached; another is that it has been associated with what look like increasingly unsustainable levels of student debt in recent years (Mitchell, 2012; Stiglitz, 2012). There is also some evidence to suggest that, in spite of the incentives for private investment existing in these regimes, overall levels of educational investment have been insufficient. Goldin and Katz (2008), for example, cite the failure of the US education system to provide an adequate supply of college-educated workers to keep pace with technological change as one of the primary causes of the increase in inequality in that country at the end of the last century, while evidence in our volume indicates that even in the highly decentralised wage-setting environments found in Liberal regimes, increases in public investment in school and college based education have significant positive effects on employment in dynamic service sectors (Wren, Fodor and Theodoropoulou, chapter 3, PET).
The countries of central and Northern Europe face a different set of challenges as regards educational policy. Here high levels of co-ordination in wage bargaining ensure much higher levels of wage equality. One effect of this, though, is to reduce the incentives for private individuals to invest in higher level skills – since the relative rewards to such investment are substantially smaller. In these regimes, then, there is a risk of a shortage in the skills on which expansion in high-end knowledge intensive services relies, unless the government steps in to subsidise them (see Ansell and Gingrich, chapter 6; Iversen and Soskice, chapter 2; and Wren, Fodor, and Theodoropoulou, chapter 3, all in PET).
In the social democratic regimes of Scandinavia, this is precisely what governments have done – providing high levels of investment in school and college based education all the way from the pre-primary to the tertiary level, which have resulted in high levels of tertiary enrolment and have facilitated the expansion of high-skilled employment in highend service sectors. This strategy has several potential advantages in terms of equity. It does not rely on the existence of wage-premia for highly skilled workers to induce investment in higher level skills. It can facilitate greater equity of access to tertiary education – in the first place because that education is publicly financed, but also because the public financing of education for school-aged, and even more critically, pre-primary children has knock-on effects on levels of equity in tertiary outcomes for children from different social backgrounds (Heckman and Jacobs, 2010). Finally, investment in early childhood education and care removes some of the costs of caring from women, increasing levels of equity between men and women in terms of access to labour markets, and facilitating women’s labour force participation and employment.
The question, of course, is whether the implied spending is affordable in an era of ‘austerity’. In a context in which higher level skills are increasingly critical to expansion in the dynamic, high-value added sectors of the economy, though, one might equally ask whether countries can afford not to undertake investments of this nature. And to the extent that educational investment forms the basis for growth and expansion in high value-added sectors, and the creation of employment in these sectors, this strategy has the potential to be economically self-sustaining.
In contrast to the Scandinavian social democratic regimes, some continental European countries – like Germany – have combined high levels of co-ordination in wage setting (and equality) with lower levels of public investment in tertiary and schools based education, and levels of tertiary enrolment are relatively low. In the past of course, as Hall and Soskice (2001) have influentially argued, this formed part of a highly effective educational strategy in which large proportions of the workforce participated in apprenticeship based vocational training regimes which equipped workers with strong firm and sector-specific skills and formed the basis for comparative advantage in core areas of industrial production (for example, capital goods).
The question, however, is whether this strategy remains sustainable in an era in which employment expansion increasingly relies on exploiting the complementarities between ICT and college educated labour. Even in Germany, the archetype of the successful apprenticeship-based political economy, the proportion of workers employed in high-skilled industrial jobs has declined sharply in recent decades.
Challenges in the political arena
The structural change engendered by the transition to services is not confined to the economic arena. Our research indicates that the transition is also associated with shifts in political demographics which have important implications for electoral outcomes in general, and for welfare state politics in particular. These changes also stem directly from the transformation of labour markets.
First, we find that the transition to services has changed the extent to which workers at different skill levels are exposed to international markets. Half a century ago, workers at lower skill levels were more likely to be employed in sectors which were exposed to international market forces than their higher skilled counterparts. This was because employment at lower skill levels was heavily concentrated in (typically internationally traded) manufacturing sectors, while employment at higher skill levels was more concentrated in (typically untraded) service professions (such as medicine, government, and the law). Our data shows, however, that one effect of the ICT revolution, and the expansion of high skilled employment in internationally traded service sectors like business and finance over the past two decades, has been to significantly increase the proportion of highly skilled workers who are employed in internationally exposed sectors (as the numbers of highly trained workers finding employment in these sectors increases). In contrast, at the lower end of the skill spectrum, the dominant effect of de-industrialisation has been a decline in the proportion of workers employed in internationally traded manufacturing sectors and an increasing concentration of employment in sheltered service sectors (non-traded personal or consumer services and the public sector).
These changes may not seem particularly noteworthy, but in fact they have important political implications. Our analysis of public opinion data indicates that highly skilled workers employed in sectors which are exposed to the global economy are significantly less supportive of welfare state spending and redistribution than their counterparts at similar skill levels working in sheltered sectors: they are also less likely to vote for left parties (Rehm and Wren, chapter 8, PET). We attribute this variation in part to a perception amongst workers in exposed sectors that levels of welfare state spending which are too high pose a threat to competitiveness and, ultimately, to their employment. Of course this (real or imagined) threat must be balanced against the benefits of welfare state spending in terms of public service provision and income guarantees. However, for workers at higher incomes, the net benefits of social policy are less likely to outweigh the cost, especially where the perceived threat to their employment and income through lost competitiveness is high, with the result that they are less favourable to welfare state spending and redistribution, and less likely to vote for parties of the left.
Highly-skilled and highly-paid workers in sheltered sectors, in contrast, are less likely to perceive that welfare state spending poses a threat to their jobs – and indeed if they are employed in the public sector itself they clearly rely on such spending for their own employment. As a result, these kinds of workers are more likely to support higher levels of welfare state provision, and more likely to vote for left parties. At lower skill levels, we find little significant difference in the preferences of workers in exposed and sheltered workers. It is well established that income is one of the primary determinants of levels of support for social policy (and unsurprising given that the net benefits of social policy typically decline as incomes increase). At lower incomes, therefore, it appears that the net benefits of social policy in terms of income and welfare guarantees outweigh the risks of employment loss even for workers in exposed sectors, so that their preferences roughly align with their sheltered counterparts.
This is important because it reduces the capacity for cross-class consensus in support of the welfare state, and left parties. Whereas previously highly-skilled workers employed in sheltered service sectors provided a critical anchor in cross-class coalitions in support of left-leaning economic platforms, the movement of significant numbers of highly skilled workers out of these sectors and into internationally exposed private sectors threatens to undermine these coalitions (since levels of support for welfare state spending and left parties are typically lower amongst these groups). Moreover, additional findings in the volume suggest that other changes in labour markets closely associated with the increased economic dominance of this kind of sector may have similar kinds of implications – for example, we find that workers who work more hours than the typical working week (common in many high end service sectors) are also less favourable to various redistributive policies and to left parties than their colleagues (at similar skill and income levels) who work shorter hours (Barnes, chapter 9, PET).
Bad news for the left and for the welfare state then? The good news, however, relates to the other dominant aspect of the transformation of labour markets in service based economies – that is, of course, the rapid increase in the number of female labour force participants. Female labour force participation closely tracks the expansion of service sector employment, and the indications are that causality in this relationship runs in both directions. On the one hand, women have a comparative advantage in service sector jobs in which they are not disadvantaged, either by a premium on ‘brawn’, or by labour market experiences which may be interrupted (since service sectors are less heavily reliant on firm and sector specific skills acquired through long-term experience within a particular firm or sector, and are reliant instead on more general skills, typically acquired initially through schools and college based learning, and more easily transferable across firms and sectors) (Iversen and Rosenbluth, 2010 and chapter 10, PET). On the other hand, increases in female labour market participation in themselves clearly increase the demand for services which might otherwise be provided in the home (such as, for example, childcare, cleaning, and catering).
Whatever its roots, the relationship between service sector expansion and female employment is a tight one. The importance of this for politics is that women who participate in the labour market are significantly more supportive of welfare state policies involving public employment and service provision. Public service expansion serves simultaneously to provide female employment and to facilitate female labour market participation (by supporting women in the caring roles for which they typically share a disproportionate amount of the burden like child and elder care). As a result, women who participate in the labour market now show higher levels of support for these policies, and for the left parties that advocate them, than men; they are also more left-leaning than women who do not participate. And as a result, the size of the gender gap in political preferences is increasing over time in line with rising rates of female labour force participation (Iversen and Rosenbluth, 2010, and chapter 10, PET).
This constitutes a significant change in gender-based patterns of political preferences, which has potentially radical implications for politics: in the past, the gender gap between men and women, if anything, went the other way, with women typically displaying more conservative preferences than men. The close association between the switch to the left in women’s political preferences, and the expansion of women’s employment in service sector jobs, however, means that this new feature of the electoral landscape is unlikely to shift in the near future.
The service transition thus creates significant new challenges for governments in both the economic and electoral arenas. On the one hand it puts pressure on existing socioeconomic models, creating new kinds of choices over policy and distributional outcomes; on the other it threatens traditional coalitions based on confluences of interest in the industrial economy, and opens up opportunities for the formation of new coalitions adapted to the dynamics of political preference formation in post-industrial labour markets.
While they face common challenges, alternative ‘varieties of capitalism’ have experienced the transition, and responded to it, in somewhat different ways.
Liberal regimes have been reasonably successful at creating employment in private service sectors. However, this employment creation has come at a cost. It has been facilitated in part by high levels of wage inequality which simultaneously generate demand for personal and consumer services by keeping relative prices low, and increase the incentives for private investment in the higher level skills required for successful expansion in high productivity knowledge-intensive sectors. It has also relied on high levels of private sector indebtedness to fuel the expansion of consumer demand for services (as well as to finance tertiary education).
The experience of these regimes during the financial crisis and subsequent recession has clearly revealed that aspects of this model are ultimately unsustainable – in particular the over-reliance on the debt fuelled expansion of domestic demand (for low end services in particular) as a mechanism for employment creation. Meanwhile evidence suggests that the partially-private model of education financing is not only associated with increasingly onerous debt burdens, but may be falling short in terms of providing the skills which expansion in more dynamic knowledge-intensive service sectors requires, hinting at the importance of public investment in education in these regimes to the development of a more sustainable service sector growth strategy. In political terms, the Liberal regimes in particular face the greatest risk of an intensification of political divisions between low and high-skilled workers, since this is where these groups are most polarised in terms of their incomes, and levels of exposure to the international economy.
The co-ordinated regimes of the central European continent, meanwhile, have kept unemployment levels low in recent decades in part because their industrial export performance has remained relatively strong, and in part because rates of labour force participation in these countries remain low (the notable exception to this profile is the Netherlands, as we discuss in the book). However, even in Germany, the industrial powerhouse, de-industrialisation is taking place, and the proportion of the working-age population currently employed in traditional sectors in Germany is not significantly greater than elsewhere.
The problem for these countries is that they have been less successful at expanding employment in service sectors. High rates of wage equality have the simultaneous effects of keeping relative wages and prices for personal and consumer services high, restricting the expansion of demand and employment, and reducing the incentives for individual investment in the higher level skills which successful competition in ICT-intensive service sectors require. In the past, low rates of tertiary enrolment in these regimes were not a problem, because they were dovetailed with the highly developed apprenticeship-based vocational training systems upon which their comparative advantage in industrial sectors relied. In the current era of ICT-intensive growth, however, restricted tertiary enrolment is likely to prove more problematic.
Further, the effective exclusion of large numbers of women from labour market participation in these regimes as a result of low levels of (public and private) childcare provision and restricted service sector employment opportunities, raises concerns first of all relating to equity, but also relating to the sustainability of the welfare state (as the core of the male industrial workforce shrinks in size, the affordability of welfare state provision must come into question in the absence of an expansion of service sector jobs and female employment). It is in these regimes in particular that the potential for radical change stemming from the increasing gender gap in political preferences seems greatest. As the number of working women creeps up in these countries, and with it the demand for public service provision and employment, one possible outcome could be a political shift to the left.
Finally, the social democratic co-ordinated regimes of Northern Europe have (to varying degrees) successfully combined employment creation with equality in recent decades by several means. First, the expansion of public service employment has served in some measure to compensate for restrictions on job creation in low-skilled service sectors (like personal services) in which the capacity for productivity growth is low. Second, public subsidisation of school and college based education all the way from the pre-primary through to tertiary levels has ensured high rates of tertiary enrolment, even though the wage premia available to highly educated workers are low in relative terms. Finally, as in Germany and other co-ordinated regimes, wage co-ordination itself has played an important role in ensuring the continued competitiveness of export sectors. These countries have performed well also in terms of gender equity, with female participation in labour markets facilitated through the creation of public sector jobs in services, and a range of welfare state policies such as the provision of state subsidised childcare.
The future sustainability of this strategy in economic terms, of course, depends on its continued affordability, and, in an era of de-industrialisation, this increasingly relies on the capacity of these countries to generate expansion in high productivity, ICT-intensive service sectors. To this end, continued high levels of investment in schools and college based education, beginning at the earliest levels, will be key. Ironically, though, successful expansion in high-end services could simultaneously pose a threat to the political sustainability of this strategy. As increasing numbers of highly skilled workers move out of sheltered welfare sectors and into high-end service sectors which are exposed to the international economy, the challenge for governments will be to maintain cross-class coalitions supportive of high levels of public provision, equality and taxation. In this context, the large cohort of working women in Scandinavian countries (who rely on these policies to facilitate their labour force participation and employment) is likely to have an increasingly critical political role to play in anchoring these coalitions.
Ansell, B. and Gingrich, J. (2013) ‘A tale of two trilemmas: varieties of higher education and the service economy’, in Wren, A. (ed.) The Political Economy of the Service Transition, Oxford, Oxford University Press.
Autor, D. H., Levy, F. and Murnane, R. J. (2003) ‘The skill content of recent technological change: an empirical exploration’, Quarterly Journal of Economics 118 (4): 1279-1333.
Barnes, L. (2013) ‘The political economy of working time and redistribution’, in Wren, A. (ed.) The Political Economy of the Service Transition, Oxford, Oxford University Press.
Cuhne, F. and Heckman, J. J. (2007) ‘The technology of skill formation’, American Economic Review 97 (2): 31-47.
Goldin, C. D. and Katz, L. F. (2008) The Race Between Education and Technology, Boston, The Belknap Press of Harvard University Press.
Gregory, M., Salverda, W. and Schettkat, R. (2007) Services and Employment: Explaining the U.S. European Gap, Princeton, Princeton University Press.
Hall, P. and Soskice, D. (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford, Oxford University Press.
Heckman, J. J. and Jacobs, B. (2010) ‘Policies to create and destroy human capital in Europe’, NBER Working Paper No. 15742.
Iversen, T. and Rosenbluth, F. (2010) Women, Work, and Politics: The Political Economy of Gender Inequality, New Haven, Yale University Press.
Iversen, T. and Soskice, D. (2013) ‘A political-institutional model of real exchange rates, competitiveness, and the division of labor’, in Wren, A. (ed.) The Political Economy of the Service Transition, Oxford, Oxford University Press.
Iversen, T. and Wren, A. (1998) ‘Equality, employment, and budgetary restraint: the trilemma of the service economy’, World Politics 50 (4): 507-74.
Meidner, R. (1974) Co-ordination and Solidarity: An Approach to Wages Policy, Stockholm, Bokforlaget Prisma.
Mitchell, J. (2012) ‘Student debt rises by 8% as college tuitions climb’, Wall Street Journal 31.5.2012.
Nelson, M. and Stephens, J. (2013) ‘The service transition and women’s employment’, in Wren, A. (ed.) The Political Economy of the Service Transition, Oxford, Oxford University Press.
Rehm, P. and Wren, A. (2013) ‘Service expansion, international exposure, and political preferences’, in Wren, A. (ed.) The Political Economy of the Service Transition, Oxford, Oxford University Press.
Rehn, G. (1985) ‘Swedish active labour market policy: retrospect and prospect’, Industrial Relations 24 (1): 62-89.
Stigliz, J. (2012) ‘Debt buries graduates’ American Dream’, USA Today Weekly International Edition 13-15.7.2012.
Wallerstein, M. (1990) ‘Centralised bargaining and wage restraint’, American Journal of Political Science 33 (4): 982-1004.
Wren, A., Fodor, M. and Theodoropoulou, S. (2013) ‘The trilemma revisited: institutions, inequality, and employment creation in an era of ICT-intensive service expansion’, in Wren, A. (ed.) The Political Economy of the Service Transition, Oxford, Oxford University Press.
Wren, A. (ed.) (2013) The Political Economy of the Service Transition, Oxford, Oxford University Press.