a literature review on skills and innovation how does successful innovation impact on the demand for skills and how do skills drive innovation

A recent study of French manufacturing firms, however, found that both product and process organisational innovation are negatively related to employment, except when they are both intro

A Literature Review on Skills and Innovation How Does Successful Innovation Impact on the Demand for

Skills and How Do Skills Drive

Chapter 3 Innovation and Its Implications for the Demand for Skills 26

Chapter 4 The Instituted Supply of Skills and Implications for Innovation 45

Chapter 5 Innovation Cycles and Strategies, and the Role of Skills 73

Chapter 6 Conclusions 96

Case Study Boxes Page

Disclaimer

The views expressed in this report are those of the authors and do not necessarily reflect the

views of the Department of Trade and Industry, or any other individual or body associated with

this research

Acknowledgements

The authors would like to thank the members of the DTI’s project steering committee - David

Campbell, Ben Chesson, Maria Cody, Samantha Hanna and especially Ben Marriott, the project

manager, for their guidance, comments and suggestions, patience and understanding We would

also like to thank Damian Grimshaw, Ian Jones, Geoff Mason and Mario Pianta for helpful

comments and suggestions

Executive Summary

The Government is concerned with productivity, which is essentially a measure of ability to create wealth The more productive people are, the more wealth they create for every hour they work It has been found that the UK has a productivity gap with some of its main competitors - including France, Germany and the US - enabling those countries to create more wealth per hour worked Attention has therefore been focused on what the UK can do to speed up its productivity growth and create more wealth from the effort its workers expend

The Government has identified five key drivers of productivity, which are innovation, skills, investment, enterprise and competition The direct links between each of these drivers and productivity have been explored in detail, but thus far relatively little attention has been focused

on how some of these drivers interrelate with each other to drive productivity

This research is intended to start to explore the links and relationships between two of the productivity drivers: skills and innovation It does this through pulling together theoretical and empirical literature from across various academic disciplines, including labour economics, management studies and innovation studies

A skill can be defined as an ability or proficiency at a task that is normally acquired through education, training and/or experience It can at times be synonymous with the related concepts

of competence, expertise, knowledge and human capital There are many different kinds of skills

In this report, distinctions will be made between different levels of skill, and different types of skill The types considered include pure science, engineering, problem-solving, language skills, team working and communication skills

Innovation is defined as the successful exploitation of new ideas That is the development and commercial exploitation of a new idea for a product or process that contributes to wealth creation and profitability The large-scale benefits of innovation come from the eventual wider diffusion of these new products and processes across the economy Innovation has many forms

It can be technological in nature (relating to new machinery or other forms of equipment), organisational (relating to changing management practices or general structures) or even new individual ways of working unrelated to colleagues or new equipment

How innovation drives the demand for skills

This report finds that the impact of innovation on demand for skills is very dependent on the type of innovation and the context Most of the literature examined focuses on technological innovation In the case of process innovation,1 the literature predicts that innovation reduces jobs, because it tends to be of the kind where capital (new machinery/equipment) replaces labour, particularly unskilled labour However, this is not always the case, as a process innovation could be of the kind where more labour is used relative to capital, or where the job losses are offset by the creation of jobs elsewhere in the company or the economy (such as in the upstream production of new machinery) The empirical literature suggests, however, that process innovation, in general, implies job losses The theoretical and empirical literature suggests that product innovation, on the other hand, is positive to employment levels as it will often result in

an increase in demand for those products

1 The development and commercial exploitation of a new way of producing the organisation’s product(s)

The report goes on to discuss what types of employment are created or destroyed by innovation, and hence the changing demand for different types of skills The labour economics empirical literature is in broad agreement that technological innovation in advanced western economies has been skill-biased That is it has increased the demand for higher skills and reduced the demand for lower skills This skill-biased technical change has particularly characterised the US in recent years, and to a lesser extent the UK Labour economists tend to agree that skill-biased technical change has contributed to altering the distribution of earnings, exacerbating wage inequalities and increasing sector-specific unemployment

Organisational innovation has been less studied but what literature there is suggests that organisational product innovation is positive to employment and that organisational process innovation is negative to employment A recent study of French manufacturing firms, however, found that both product and process organisational innovation are negatively related to employment, except when they are both introduced in combination and aimed at increasing market size, in which case they are positively related to employment

In assessing the effect of innovation on the demand for skills, it is worth remembering that innovation is only one of a series of factors affecting the demand for skills Other factors include labour market institutions, such as the minimum wage and trade unions, domestic changes in the conditions of competition, the composition of public spending and changes in workers and consumers’ preferences The complex interactions between these factors make estimations of their interdependent effects extremely difficult International trade is another important factor that drives the demand for skills It is often framed as an alternative explanation to innovation in changing the demand for skills, but, as researchers have realised increasingly, it can often be complementary (reinforcing) to the effects of innovation

How the supply of skills drives innovation performance

There is no one mix of skills that is conducive to good innovation performance in all circumstances Instead, the required skills vary across the type of innovation concerned, the industry and the strategic model the firm pursues

In the older models of (technological) innovation, such as the ‘science push’ model of Vannevar Bush, the skills to ‘produce’ innovation are the degree and higher-level science and engineering skills of a small head or elite in the organisational hierarchy More recent models of innovation, such as the ‘systemic integration’ model, allow for more democratic, distributed sources of innovation, involving the skills of the whole workforce In particular, this model says all workers should have basic ‘platform’ skills that allow them to be adaptable to changing circumstances and more open to new ideas to be innovative This contrasts with the past, when many workers developed only job-specific, narrow skillsets This model emphasises the importance of interaction and cooperation between all the actors in the innovation process, including the innovating firms’ suppliers, commercial partners and customers These interactions require communicating and negotiating skills

The report identifies management and leadership skills as being of particular importance for all types of innovation Without leadership, or ‘strategic intent’, firms do not innovate in any fundamental ways, and have to rely on small incremental improvements to their ways of working from learning-by-doing In a competitive environment, managers have to choose between different innovation and growth strategies, even if they are just to keep ‘treading water’ Putting the chosen strategy into practice requires management and leadership skills at all levels of the organisation There is little empirical evidence about how management skills (which may be

imperfectly measured in a variety of ways) relate to innovation, but theoretically they are thought

to be very important

Intermediate level technical skills are also thought to be very important for innovation, especially

in manufacturing This is significant because the UK has a smaller proportion of people with these skills than countries like Germany and Japan Further, there is a concern that the quality of these skills is relatively low in the UK because of an over-reliance on narrow job related skills rather than underpinning knowledge (transferable skills) The German ‘dual’ apprenticeship system teaches apprentices academic (in the sense of conceptual or theoretical understanding) as well as vocational skills This is thought to enhance the capacity of German workers to adapt to, and enhance the new technologies they encounter over their working lives Evidence about the effect of these skill differences comes from the NIESR matched-plant studies, which show that British managers tend to defer the implementation of new equipment until it is ‘bug-free’, in part because of acknowledged skills weaknesses on the shop floor These delays imply British manufacturing firms cannot build up the valuable relationships with leading-edge suppliers, unlike German firms They therefore end up competing on price, and not on quality, and are considerably less productive

Other (mixed) evidence of the importance of the supply of skills for innovation performance comes from surveys asking about the effects of skills shortages The UK Innovation Survey (relating to 1998-2000), suggested that lack of qualified personnel was not a first-order barrier to innovation, coming fifth out of nine barriers However, in the European Innobarometer Survey (2001), finding and mobilising human resources was the most cited barrier to innovation in the

UK More recent data from the National Employers Skills Survey 2003 found that over a fifth of the firms that reported skills gaps amongst their existing workforces had delayed implementing new products due to these deficiencies, whilst nearly a third had experienced difficulties introducing new working practices It is worth mentioning that skills is one of several factors affecting innovation performance Other important barriers include access to finance and lack of demand for innovative products

Since the supply of skills is less than socially optimal, the report asks why firms and workers do not invest more in education and training Becker’s theory of human capital suggests that there may be several reasons, including lack of information on the returns to education and training, lack of access to finance to fund it and, for firms, fear of not reaping the return on their investment in training, because their trained workers may leave for another company The latter problem, fear of so-called poaching,2 may be especially relevant to highly-innovative, high-technology industries In this case, the training may be extremely expensive, risky, yet very valuable to the firm that reaps the return on the investment, especially if a new market is being opened up

The dynamic relationship between skills and innovation

The questions about how innovation drives demand for skills and which skills are required for innovation are to some extent two sides of the same coin It is difficult to disentangle the skills that drive innovation from those which are demanded as a result of change brought about by innovation In the long term, the relationship between skills and innovation must be circular The skills of the workforce and management will help determine the innovation that takes place, which will then help determine the changed demand for skills in the firm, which will influence the innovation that takes place and so on

2 There is evidence that fear of poaching is more of a problem than actual poaching

The product life cycle theory provides some answers to how innovation changes the demand for skills through the stages of a new product innovation, or it could also be viewed as how the skills for successful innovation change over the cycle.3 In the early, fluid stages of the industry where the product is ill-defined, the key skills are those of entrepreneurs, and sometimes those of scientific or technical specialists, which together with marketing skills realise or create the market for the product As time goes on, the product tends to become more standardised, until a dominant design emerges In this transitional stage, there is a shift from product to process innovation Functional, scientific management skills and specialist workforce skills are increasingly required Once the dominant design is established, the industry enters the specific stage, where innovation is more incremental and cumulative The skills required are managerial

‘command and control skills’ and low level or unspecific workforce skills Higher-level workforce skills may be increasingly required if the firm attempts to move into the higher quality end of the market in response to low-cost competition The product life cycle model had been principally applied to manufacturing firms There is limited evidence that some service sectors move in the opposite direction, from large-scale provision of standard services and process innovation to smaller-scale provision of customer-specific services, implying product innovation

Low skill equilibrium is a relevant concept in the literature relating to the interaction of supply and demand for skills and innovation Essentially, this is a situation in which firms encounter a demand for low-specification, low-value added products, which only require low skills to produce They therefore tend to compete on price, utilising low-skill workers As low-skill jobs are available, but relatively few high-skill jobs exist, workers have little incentive to acquire higher-level skills, and so firms do not have the supply of higher-level skills that they would require if they wanted to move to a high-specification product strategy This situation is thought

to pertain in some parts of the UK economy This theory is supported by the empirical work of Geoff Mason who suggests that demand for skills in Britain is lower than that of several other economies because a large proportion of British enterprises have adopted low specification product strategies Further he argues that in the absence of strong incentives like foreign competition there is no reason for firms to change and adopt an innovation based strategy Also when firms wish to move up market they need different skills and so are likely to then experience skill shortages

Another aspect that affects the relationship between innovation and skills is the industry in which the firms operate In the innovation studies literature, Pavitt argued there are four different types

of (mainly manufacturing) industry that have their own patterns of innovation and requirements for skillsets First, science-based industries such as pharmaceuticals and electronics, require a core

of degree-level science and engineering skills, and work a lot with universities and/or have their own R&D laboratories Second, scale-intensive industries rely on economies of scale, exploited

by scientific managers, increasingly with cross-functional integration skills, specialist product design, engineering and development skills, and a workforce that is able to adapt to and make the best use of technologies Third, specialist suppliers, which engage in a lot of cooperation with lead users, tend to require the skills of interactive learning, the expertise to develop highly client-specific solutions and vocational, practical development skills Fourth, supplier-dominated industries, where most (technological) innovation comes from outside the firm, require entrepreneurial skills for spotting market trends, and a workforce with the ability to implement and use new technologies

In a later paper, Pavitt added a fifth category of information-intensive firms, largely in the service sector Their main source of new technology is in-house software or systems development,

3 Not all industries will follow this pattern, as has been shown empirically

although some may come from outside Here the style of innovation requires technical skills to develop software and communication or cooperation skills to acquire technology from outside The empirical evidence suggests that service-sector firms are more organisationally innovative than manufacturing firms These organisationally-orientated service-sector firms are more likely

to emphasise the importance of supply chain interaction and external intellectual property but less likely to emphasise in-house R&D and research based cooperation with other organisations such as universities as their key sources of technology

Much of the literature emphasises the systemic nature of the relationships between innovation, skills and productivity Florida proposes the concept of the ‘creative class’ of people – scientists, engineers, designers and creative arts workers He finds their concentration level in local populations is positively related to economic performance and tolerance Meanwhile, in the innovation studies literature, Kodama suggests that with technological fusion and increasing cooperation, innovations are increasingly crossing industry and technological boundaries, which will tend to make the new innovations more widely applicable This in turn is changing the demand for skills by requiring that workers have broader, less specialised skills, involving multiple disciplines

Policy implications

With the growth of trade in goods and services, and the growing internationalisation of scientific and technological activities, it is increasingly important that innovation in the UK involves a large proportion of firms and their employees No longer can companies or the country as a whole rely heavily on small elites of highly trained or educated people Instead, innovation needs to be organisational as well as technological, and to involve those on the shop floor as well as those in R&D, marketing and management positions It follows, therefore, that one of the over-riding points is that if the UK is to continue to have a dynamic innovative economy then all its people need to have a good general education This needs to provide the person who receives it with a platform of generic skills that will prove useful in engaging with and adapting to the changing circumstances that will doubtless pertain within and across industries and firms That is they will

be better able to innovate and adapt to the effects of innovation in their own company, and better able to re-enter the labour market, even moving into different fields, if they are made redundant

The report suggests that a certain amount of academic education alongside vocational education raises the individual’s absorptive capacity to learn and adapt And, again, innovation is likely to require an increasingly adaptive workforce with a broad range of skills

The supply of skills seems to be less than would be socially optimal, even with the current Government interventions in this area Therefore, perhaps some new form of cost and benefit sharing is required between the employer, employee and the Government, to encourage greater skill formation The authors explore the possibility of tax credits for training, as have been introduced in Canada An argument against such an approach is that it would be difficult to ensure that tax credits were not just being given for training that would have occurred anyway Another issue that needs to be addressed is the skills and commitment of managers to innovation and the R&D tax credit may prove to be a valuable step in this direction Changing routines and ingrained practices is, of course, difficult Encouraging more collaborative, trust-based arrangements between firms is necessary whereby firms may gain long-term advantages from interactive learning, even if there are additional short-run costs This is best achieved by confronting firms not with a single ‘best practice’ model, but rather by encouraging step-by-step

transformations through the adoption of practices appropriate to their stage or level of commitment to innovation

There perhaps should also be more consideration of the trade-off between short-run productivity, or static efficiency, and capacity to innovate, or dynamic efficiency In many companies and organisations a battery of targets or financial measures encourage short-run efficiency but fail to recognise the importance of ‘organisational slack’ and local experimentation

to achieve innovation and greater dynamic efficiency 3M allows its engineers to spend 15% of their time on any project that happens to interest them This approach, while clearly not applicable to all, has been shown to boost innovation, as well as increasing employee commitment and motivation

1 Introduction

1.1 The DTI’s definition of innovation is the successful exploitation of new ideas That is

the development and commercial exploitation of a new idea for a product or process that contributes to wealth creation and profitability The large-scale benefits of innovation come from the eventual wider diffusion of these new products and processes across the economy Innovation takes many forms It can be technological in nature (relating to new machinery or other form of equipment), organisational (relating to changing management practices or general structures) or even outside of these two main categories

A skill can be defined as an ability or proficiency at a task that is normally acquired or developed through education, training and/or experience It is, of course, related to the concepts of competence (which tends to imply these days that there is a demand for the skill), expertise and knowledge There are many different kinds of skills

1.2 The aim of this study is to review the literature in order to start answering the questions:

- Is there a mix of skills at either the firm level or within society that would increase the amount of innovation that is undertaken?

- How does the successful introduction of new ideas change the pattern of demand for skills in companies and the economy?

- How do these two questions relate to each other and what is the nature of the way relationship between skills and innovation?

two What are the policy recommendations that arise?

- What are the gaps in the evidence base that require further research?

1.3 To address these questions we have undertaken a review of the literature, or more

specifically a review of three strands of literature, each of which is discussed in a separate chapter

1.4 The report is structured as follows Chapter 2 sets the scene regarding skills, innovation

and wider economic performance in the UK This is intended to set the study in its wider

context Chapter 3 is largely based on the mainstream economics literature and

essentially concerns how innovation, particularly in the form of ‘technical change’ is impacting on the demand for skills in the economy The literature is essentially in agreement that, for an advanced economy like the UK, innovation is positively skill

biased, although there is a debate about why this is the case Chapter 4 is largely based

on labour process studies on the supply of labour and how the instituted processes of skill formation provide countries with different endowments in terms of skills, which then impact upon countries’ abilities to engage in innovation, of different forms

Chapter 5 is based on an understanding of innovation and skill requirements from an

innovation studies / Schumpeterian economics perspective.4 This approach is more systemic, and is not based around the classic supply and demand framework of mainstream economics

4 Schumpeter (1883-1950) was an Austrian economist who placed innovation, and the ‘creative destruction’ of innovation at the heart of his analysis Today, Schumpeter's views are especially influential amongst heterodox and evolutionary economists, and innovation management scholars, who are interested in how innovation shapes competitiveness, industrial organisation and economic development Schumpeterian ideas have also been highly influential in shaping innovation policy For example, the EU’s innovation program, and its main development plan, the Lisbon Strategy are arguably based on Schumpeter (See http://en.wikipedia.org/wiki/Schumpeter)

1.5 In Chapter 6 we conclude the study, reviewing what we have learnt to answer the

questions as outlined above

2 Context: The UK Economy’s Skills & Innovation Performance

Introduction

2.1 Overall, relative to its historical performance, the UK economy has enjoyed a decade of

relatively strong and stable growth in GDP, with stable and low inflation, and falling unemployment.5 With inflation and unemployment under control, the chief concern has been to establish the framework conditions to stimulate the development of the economy towards higher value added activities and ‘the knowledge economy’

2.2 To achieve its ambitions, the Government recognises the UK will have to improve its

productivity and innovation performance: “We live in a fiercely competitive global economy If we are to enhance our productivity and raise our standards of living we need

to improve radically our innovation performance And we will need to innovate continuously in the future so that we can hold our own against fast moving new economies” (DTI’s Innovation Report, 2003, p 16) It is argued that UK companies will have to compete more on quality, with high value added and more innovative products and processes This, in turn, “will require inspirational leadership, stronger management and leadership skills, a highly skilled workforce, a flexible labour market that promotes

diversity and fair treatment, and high performance workplaces” (op cit, p 11) Moreover,

the Government seeks: “A highly educated workforce with a culture of lifelong learning

[which] is more likely to adapt to economic change” (ibid)

2.3 The purpose of this chapter is to set the context for the studies examined in the report in

the light of the government’s objectives and the key structural characteristics of the UK economy We will sketch the macro-economic context of productivity, patterns of structural change, skills and educational attainment and innovation performance

Productivity and Value Added

2.4 Productivity is in essence a measure of the efficiency of an economy in terms of the

amount of output produced per unit of input used Relatively simple measures of productivity include output per worker, or output per hour worked More complex measures include total (or multi-) factor productivity which takes into account not only labour inputs but also capital (buildings, machinery, equipment, etc.), the stock of which

is estimated rather than measured precisely The UK Government’s preferred measure is output per worker, which has the advantage of being the most straightforward to measure and also being immediately linked to the overall objective of raising the growth trend of the UK economy The Government considers that productivity is driven by five factors: investments in capital (buildings, machinery, equipment, etc.), skills (i.e investments in people through education and training), innovation (i.e the generation and application of new knowledge), enterprise or entrepreneurship (i.e a willingness to start new businesses, take risks, etc.) and competition between providers of goods and services These drivers interact in complex ways For example, the extent of competition is a key driver of innovation (e.g a monopolist has little or no incentive to innovate), whilst innovation also impacts on competition (e.g by creating or breaking a monopoly) In this report we will discuss especially the inter-relationship between skills and innovation, within the context of raising value added and productivity

2.5 At the level of the firm, productivity can be measured by dividing value added (the value

of the outputs less the value of the raw materials and intermediate goods and services

5 In contrast, the major problems with the UK economy since the post-war boom until the mid-1980s have been inflation and unemployment (Nickell and van Reenen, 2001)

used in the production of these outputs) by the inputs used, such as the number of employees, the number of hours worked, or a measure which combines both the labour and capital used in the production process

2.6 Measures of productivity allow comparisons between countries and firms in terms of

their performance, but the current performance of a country or company is determined

to a significant extent by decisions made some years ago Similarly, decisions made today will impact on productivity in the future For example, current investments in R&D and training are being made in the expectation that the firm will gain a greater return in the long run than it would save in the short run by not engaging in these activities However, even this is complex, as many if not most firms do not have a precise understanding of the rates of return to their R&D and training activities Suffice to say that particularly in the past UK firms were often criticised for being short-termist – that is seeing to maximise their short run performance and failing to invest significantly for the longer term through activities such as R&D and training

2.7 At the aggregate level the UK has long suffered a productivity deficit relative to its major

competitor nations In terms of labour productivity, the UK lags significantly behind the

US As van Reenan (2004, p.4) puts it: “Joe Doe, our average American worker, could take Thursday and Friday off and produce as much as poor John Bull toiling away throughout the working week” Although the higher productivity of US workers means they could work fewer hours to achieve the same standard of living as UK workers, in fact they tend to work longer hours and therefore tend to have significantly higher incomes Growth accounting suggests that US productivity is higher because US workers have more and better equipment (although they have similar skills), and because of greater past investments in R&D But some of the US’s advantage remains unexplained

As Van Reenan (2004) asks: “Why do US workers create so much more output for their fixed and human capital inputs? What is it that they do more effectively that makes such

a significant difference on the macroeconomic level?” Increasing the amount of physical capital per hour worked in the UK to American levels would only make up about half the gap in labour productivity between the UK and US Nor would having the same skill distribution as the US close the gap Around half the gap between the UK and US labour productivity is effectively unexplained and is attributed to innovation and different ways

of working

2.8 In recent years, the US has enjoyed a period of rapidly rising productivity US

productivity growth rose from about 1.2% per annum between the late 1970s and 1995,

to 2.2% per annum between 1995 and 2001 Most economists reason that the diffusion

of Information Communication Technologies (ICTs) and associated innovations in working practices is an important part of the US’s faster productivity growth since the late 1990s (see Chapter 3 of this report), but there is a debate about whether the impact

of ICTs has been found mainly in the ICT producing sectors (such as software and conductor manufacturing), or also in the ICT producing and using sectors (such as retailing and transportation).6 Economists are however struggling to explain why the uptake of ICTs and new business practices associated with these has been slower in Europe than the US: “Since ICT is available at about the same price in Europe as in the

semi-US it is not clear why Europe has lagged behind So as the ICT revolution introduced new forms of capital, European economies failed to exploit its benefits in terms of

6 Among the alternative explanations for the rise in US productivity is the growth in the number of illegal workers, for example at retailers like Wal-Mart, whose input is not officially recorded but whose output is effectively attributed to the legal component of the workforce

enhanced productivity and prosperity as fully as the US” (Van Reenen, 2004).7 It could

be that these new business models and benefits from past investments in ICTs will soon diffuse to Europe, and the fact that the UK has made relatively high investments in ICTs means it should be well placed to take advantage of these

2.9 The UK’s productivity performance relative to its major European competitors, and

France and Germany in particular, is more contentious Simply put, the average French and German worker is more productive than the average British worker, but fewer French and Germans work, and those who do work fewer hours Assuming that, in all of these countries, the more productive people in the potential labour market tend to work, whilst the least productive tend to be economically inactive, it is less clear that the difference in productivity between Britain, France and Germany would still remain if the same proportion of the labour force in each of these countries worked, and especially if the average Briton, French and German worker all worked for the same number of hours and had the same amount and quality of equipment available to them

2.10 Arguably, the UK has lower productivity because it has simpler and less complicated

labour market regulations than France, Germany and most other European countries This has the direct effect of encouraging more people to work and encouraging employers to offer more jobs, not least because it is also relatively easy to reduce employment by cutting jobs This means more marginal workers are more likely to be employed in the UK, reducing unemployment and inactivity But the higher labour costs and stricter labour market regulations in Continental Europe also encourage companies

to substitute capital goods for labour, meaning that the average European worker has better equipment (i.e more and/or newer equipment) with which to work, which further raises output per person employed, resulting in higher productivity amongst those in work One study estimates that this deficit in physical capital accounts for about 80% of the labour productivity gap between the UK and France and Germany (van Reenan, 2004) The other 20% is very largely attributed to the lower skill levels of British workers (which itself partly reflects the UK’s higher employment rates)

2.11 The UK therefore enjoys higher employment rates than France and Germany, to some

extent at the cost of lower average productivity amongst those employed However, this trade-off between the employment rate and productivity amongst those employed does not explain how some smaller European countries (such as Switzerland, Denmark and the Netherlands) are able to achieve higher productivity whilst maintaining high participation rates

2.12 The UK is home to many excellent companies which achieve high levels of value added

(and high levels of value added per employee), as is shown by the DTI’s Value Added Scoreboard.8 But there is also wide variation in performance between best and worst performing businesses in the UK According to Haskel and Martin (2001) the best

7 The reasons why uptake of ICTs might have been slower in the UK and Europe than in the US include: the US being closer to a single market than Europe, which means there are greater economies of scale; the physical structure of Europe and the US are very different, and this is important in ICT using sectors such as retailing In the UK and Europe it is difficult to build new out-of-town shopping complexes which are generally more productive and achieve higher value added than generally small shops in historical and congested town centres Differences may also exist on the demand side US consumers have more readily accepted new and more efficient means of distribution, such as Amazon.com and eBay Some of these differences will be overcome over time, whilst others (such as the different physical infrastructures of town and cities) are effectively ‘locked in’

8 See http://www.innovation.gov.uk/value_added/keypoints.asp?p=keypoints#

manufacturing establishments are about five times more productive than the worst Such

variation is also likely to exist in other countries, although the extent of variation may be

greater or less The existence of this ‘long tail’ probably relates to the diffusion of

technologies, managerial practices and skills Notably, Haskel and Martin (2001) also find

that even if the lower performers were to raise their productivity to the median this

would still not close the productivity gap between the UK and its major competitors

Structural Change

2.13 Measures of productivity permit comparisons between the performance of countries and

companies, and it is nearly always possible to find ways of increasing productivity in

existing tasks to maximise short-run efficiency But in the longer run, such measures

may not be sufficient, as countries and companies that are highly developed in one

context may find it difficult to adapt to change Core competences become core rigidities

(Leonard Barton, 1992) ‘Structural change’ is associated with more fundamental

changes in types of work, types of output (e.g moving from goods to services, and mass

production to customisation) and ways of working (e.g from integrated companies to

more distributed forms of production involving greater outsourcing)

2.14 The UK has undergone tremendous economic change in the last quarter of a century

This has seen the decline of some industries and the growth of others, as shown in table

2.1 Scase (1999, p 23) observes that “more people now work in Indian restaurants than

in shipbuilding, steel manufacturing and in coal mining combined There are three times

as many public relations consultants as coal miners.”

Table 2.1 Employment * in the UK by Industry, Spring 1984 and Spring 2004 9

1984 2004 1984 2004 1984 2004 All in Employment (000s) 23,974 28,311 14,039 15,296 9,935 13,015

Agriculture & Fishing (%) 2.6 1.3 3.5 1.9 1.3 0.6

2.15 Overall, the sectors that are growing in terms of their employment are ‘business and

miscellaneous services’, ‘non-marketed services’ (such as health and education), and

‘distribution and transport’, whilst ‘construction’ is stable, and the ‘primary sector’ (i.e

agriculture and mining), the ‘utilities’ and manufacturing are in decline Whilst this is

particularly associated with the increasing participation of women in the labour force

(almost 90% of women who work do so in service sectors), male employment is also

9 Drawn from the Labour Force Survey (LFS) Historical Quarterly Supplement Table 22

changing In 1984 half of all men in employment worked in service sectors, now that

proportion is two-thirds

2.16 Meanwhile, the types of work that are growing are ‘managers and senior officials’,

‘professionals’, ‘associated professional and technical’, ‘personal and protective services’,

and ‘sales and customer services’, whilst those types of work in decline are ‘administrative

and secretarial’, ‘craft and skilled manual’, ‘factory operatives and unskilled manual

occupations’ and ‘elementary occupations’ as shown below in table 2.2

Table 2.2 Employment * in the UK by Occupation, 1971 to Spring 2004

Personal Service Occupations (%) 3 4 5 6 7.8 2.3 14.1

Sales and Customer Services (%) 5 6 6 7 8.2 4.7 12.1

Process Plant and Machine

Elementary Occupations (%) 17 18 15 14 11.8 11.9 11.6

Data for 1971-1998 from Green, 2005 Data for 2004 from the Labour Force Survey

* includes Self-Employment

2.17 This combination of changes in occupational and industrial structure has been associated

with the growth of services and the dis-integration of production (i.e the growth of

outsourcing and the decline of the vertically integrated firm) The decline of large scale

operations, and the growth of small firms and self-employment has been associated with

a growth in small business management and administrative work, relative to production

workers The growth of services and small firms attracts widely differing views with

regard to job quality and skills Some commentators are very negative Scase, for

example, observes that, “inherent in the growth of a service and information economy is

the creation of jobs that are low paid, insecure and offer limited career opportunities

Changes in the retailing sector – the decline of traditional, independently-owned shops

and the growth of supermarket chains – have created low paid and low skill jobs The

growing need for care assistants to care for an ageing population has led to the creation

of part-time, low paid employment Moreover the growing use of ICTs is generating

similar low paid jobs in call centres These are expected to make up 5 per cent of the

labour force by 2010” (Scase, 1999, p 24) Yet the growth of services is also associated

with an expansion of high-skill-high-wage work – on average managers and senior

officials earn twice as much as process, plant and machine operatives, and on average

those working in financial services earn a third more than those in manufacturing

2.18 These contrasting trends are associated with a growing diversity of jobs by quality in the

UK Goos and Manning (2003, p 77) observe that “Whichever way you look at it, there

is growing polarisation of jobs in the UK: there are more good ‘MacJobs’ and more bad

‘McJobs’ The data show there have been strong increases in the number of high paid

jobs but also significant increases in the number of low paid jobs over the last 25 years

Craft and clerical occupations in non-service industries are disappearing while the importance of low and high paid service jobs has increased” Also notable is that evidence of job polarisation is even more marked for men than for women This increased job polarisation, which is to a significant extent due to the nature of the skills British workers develop through education and training, is discussed further in chapter 4

Growing Inactivity amongst Working Age Men

2.19 As mentioned earlier, in terms of employment and unemployment, the UK labour market

has been remarkably healthy (Nickell and Quintini, 2002).10 But while unemployment is relatively low, the inactivity rate11 in the UK has not changed markedly from about 20% for twenty-five years, and what is remarkable is the convergence in the inactivity rates of men and women “Since 1975, the percentage of non-student men of working age who are inactive has risen more than five times [from 2.6% to 13.2% in 1998], around a 10 percentage point increase By contrast, this is almost offset by a nearly 10 percentage point decline in the inactivity rate of women [from 36.5% in 1975 to 26.9% in 1998]”(Nickell and Quintini, 2002, p 211) The growth in economic inactivity has been particularly striking amongst unskilled men: “Most extraordinary is the fact that the inactivity rate amongst men of working age without qualifications was 30% in 2000, compared with less than 4% some 20 years before This despite the fact that in 2000, the

UK labour market was booming” (Nickell and Quintini, 2002, p 212) Gregg and Wadsworth (2003) show that two in five men of working age (excluding students) without qualifications are now economically inactive or unemployed, with the majority being economically inactive; see table 2.3 below This lack of work, and growing exclusion from work, has been most marked in some northern cities, such as Newcastle-Gateshead, Liverpool and Glasgow.12

10 Nickell and Quintini (2002, p 203) observe that “Since the disasters of the early 1980s, UK unemployment has moved in parallel with the best performers in Europe (Denmark, Ireland, the Netherlands).” What is especially remarkable is that unemployment has been able to fall without any significant inflationary pressure Nickell and Quintini discuss the reasons for this

11 The inactivity rate refers to the total number of individuals of working age who are not students and who are neither working nor unemployed, as a proportion of the non-student working population of working age

12 “By 2002, only 25% of less skilled men [i.e the bottom 30% by qualifications] living in social housing were in work in [these] depressed urban conurbations, sharply down from [32% in] 1993 In contrast, the employment rate for similar groups in the South East was around 65%” (Gregg and Wadsworth, 2003, p 91) According the Office for National Statistics, Newham in London had, at 42.1%, the highest inactivity rate of all local authorities

in Great Britain in 2004 (see also Anyadike-Danes, 2004)

Table 2.3 Employment, Unemployment and Inactivity in the UK, 1993 and 2003*

Amongst those with No Qualifications

Amongst those with Low Qualifications #

* Working age population excluding students

# The 30% of the population which, for their age cohort, have the lowest qualifications

Data source: Gregg and Wadsworth, 2003, Table 6.1 (Labour Force Survey Data)

2.20 The growth in inactivity amongst men almost certainly reflects structural changes in the

economy which are associated with an overall shift in demand towards higher skills, and a

shift away from physical labour to more cognitive work and ‘emotional labour’ (see

Chapter 4) The fundamental problem is that there is a large body of individuals of

working age who, because of lack of skills, do not command a high enough wage in the

labour market to provide a decent standard of living for themselves and their dependents

This problem is particularly severe in Britain because the pool of very low-skill workers is

much larger than in western Europe as a whole (Nickell and Quintini, 2002, p 215)

Moreover, not only is it larger; it is also geographically more concentrated especially in

the old industrial areas where the ‘new jobs’ created to replace those lost in heavy

industry tend to be low skill and low wage, but also increasingly ‘feminine’ (Danson,

2005) This has encouraged the expansion of women’s participation in the labour market,

whilst unskilled men have found it difficult to find or take up new forms of work. 13

2.21 By contrast, those with skills and especially degrees have prospered: “in 1980, men with

degrees earned around 63% more than those without qualifications [other things being

equal] By the mid-1990s this had risen to 93% - this despite the fact that the percentage

of employees with degrees had almost doubled over the same period” (Nickell and

Quintini, 2002, p 212)

Skills and Educational Attainment

2.22 At a broad level a skill is a human ability coupled with an actual or potential demand for

that ability Narrower definitions consider that a skill is a special ability, often acquired

through specialist training.14 Skills are directly related to employment, employability and

13 Nickell (2004) notes that the growth in women who work has been greatest amongst married women whose

partner works Meanwhile, the rise in inactivity amongst men has been concentrated on married men whose

partners do not (or cease) work and amongst single men

14 Thus ‘unskilled work’ is skilled under a broad definition but not under a narrow one A good example is ironing

of clothes, which requires good hand-eye co-ordination and is very difficult to automate, yet most people can

iron and it requires little to no training In the past, ‘skills’ have been associated with specialist abilities, especially

amongst those with vocational rather than academic qualifications (i.e those completing apprenticeships), while

currently broader definitions of skills are being favoured

productivity They are also indirectly related to productivity through innovation, as we will argue in this report

2.23 Although the situation is now improving, the UK has long struggled to provide a good

basic education for all For example, according to Leach (2002), around 7 million adults

in the UK are functionally illiterate, with one in five adults unable to locate the pages for plumbers in the Yellow Pages Nearly half of adults have numeracy skills below those expected for an 11 year-old One in four adults cannot calculate the change they would get from £2 when they buy one item for 68p and two more for 45p each A 1996 survey

by the Institute of Directors revealed that four in five company directors were concerned about the basic numeracy and literacy of job applicants

2.24 International comparisons are difficult, but the Moser (1999) report found that 23% of

British adults were at the lowest levels of literacy (c.f., 12% in Germany) and numeracy (7% in Germany), and Leach (2002) claims: “According to many international surveys, Britain’s educational standards are poor compared with other countries On the most respected international surveys, including the International Adult Literacy Survey (IALS) and the Third International Mathematical and Science Study (TIMSS) the UK has performed relatively poorly, compared to the Far East and much of Europe This said, in the last round of OECD studies the UK appears to show much improved standards, and appears to outperform other countries which are widely thought to have superior education systems (see Prais, 2003 & 2004, vs Adams, 2004)

2.25 At least as proxied by qualifications, the skills in the UK labour force have been

transformed in the last 30 years In 1974 over half of all men and two thirds of women (aged 16-69) had no qualifications at all Now these figures are 15% and 19% respectively Meanwhile, in 1974 only 4% of the male workforce and 1% of the female workforce had degrees, whist 16% and 13% respectively now do (Glennerster, 2002) In 2002/03 53% of pupils in the UK gained five or more GCSEs (or equivalent) at grades A* to C, compared with 46% in 1995/96 Girls outperformed boys, with less than half

of boys achieving this standard Meanwhile, the proportion of pupils gaining two or more GCE A levels (or equivalent) has increased dramatically from 19% in 1992/93 to 39% in 2002/03 Again girls outperform boys, with 43% of girls achieving this compared with a third of boys (Social Trends, 2005). 15

2.26 There have therefore been significant improvements, but is the glass half full or half

empty? The consensus seems to be that in the future most jobs will require higher skills One estimate reported in the media is that in ten years two-thirds of jobs will require A-levels or higher education qualifications.16 It is very difficult to assess the accuracy of these claims Partially it will depend on the extent to which more routine forms of work resist or succumb to displacement through off-shoring and automation, and partially it will depend on whether new applications are found for the relatively low skilled within the workforce For example, the domestic cleaning sector may expand enormously in the next decade, as it becomes more socially acceptable both to be and have a cleaner, and as the relatively affluent decide to ‘buy time’ (for more paid work and leisure) by outsourcing more of their domestic chores But it would appear that the UK cannot afford a high failure rate from its education system, and it is worrying that around a third

15 Also controversial is whether or not qualification standards have remained constant over time Leach (2002), for example, claims standards have declined

16 Reported in the Guardian and quoted by Ruth Kelly, Secretary of State for Education http://education.guardian.co.uk/higher/careers/story/0,9856,1445788,00.html

of pupils are failing to achieve the desired Key Stage-3 standards in English, Maths and

Science, as shown in table 2.4 below

Table 2.4 Pupils in England reaching or Exceeding Expected

‘Key-Stage 3’ (i.e at age 11-12) Standards in English, Maths and Science

reflects a number of developments.17 Partly, it reflects a social inclusion and

‘employability’ agenda, with the ability to undertake basic tasks being seen as crucial to

the employment of the unqualified, including many of the inactive and long term

unemployed Partly it also reflects the development of a qualifications culture, such that

any abilities, no matter how basic, are increasingly subject to qualifications.18

2.28 International evidence shows that in virtually all OECD countries participation in tertiary

education is increasing – i.e., young adults are more likely that the rest of the workforce

to have participated in tertiary education (see Figure 2.1) The OECD data also shows

that the UK’s performance in terms of participation in tertiary education is moderate,

with the proportion of young adults in the UK who have participated in tertiary

education being higher than some countries (notably Germany, Switzerland, the

Netherlands and Italy), but lower than in others (including Canada, Japan, Korea, the US,

Sweden and France)

17 ‘Basic skills’ are defined by the Basic Skills Agency as “the ability to read, write and speak in English and to use

mathematics at a level necessary to function and progress at work and in society in general”

18 See the National Qualifications Framework (available at http://www.qca.org.uk/493.html)

Figure 2.1 Tertiary Education - Attainment levels by 2002 (OECD Data)

d

SwedBelgiumSp

ain Ire

d Fr ce

Australia

Unite

d King

domDe ark

New

Zeala Iceland Ne rla s

Germ

anyM

ico Po d Po

Czec

h Replic Slova

Public Expenditures Private Expenditures

2.29 Certainly if the UK wants to be at the forefront of the knowledge economy it will have to

continue to address some of its deep-seated problems and invest heavily in education The Labour Party Manifesto of 2005 declares that ‘Education is our number one priority’ and claims that state funding for education has risen from 4.7% of GDP in 1997 to 5.5%

in 2005 This level of spending is still not exceptionally high by international standards as Figure 2.2 shows Of course, it is not just a matter of spending, but also how the money

is used (Leach, 2002)

The UK’s Innovation Performance

2.30 We now turn to innovation, which is “… a major determinant of productivity

performance and … differences in innovation performance are a significant cause of the UK’s relatively weaker productivity performance” (DTI, Competing in the Global Economy – the Innovation Challenge, p 9) The DTI defines innovation as: ‘the successful exploitation of new ideas’ These ideas must be new to the firm, and may also

be new to the world, the UK, or an industry

2.31 As with skills, economists have tended to make the most of the available statistics, most

notably on research and development (R&D) activities and patenting, even though these are severely biased towards technological forms of innovation (as opposed to organisational forms of innovation).19 The recent development of the European Community Innovation Surveys (CIS) has broadened the available evidence on innovation, and innovation performance, although these surveys retain a strong technological bias Here we provide a brief summary of the UK’s relative performance with respect to R&D activities and patenting, before summarising the evidence from the second and third European Community Innovation Surveys on how the UK performs in comparison with other European countries

2.32 The UK’s Innovation Performance – R&D UK business expenditure on R&D

declined through the 1980s until the late 1990s (from 1.5% of GDP in 1981 to 1.16% in 1997) The UK is however seeking to raise R&D as a proportion of GDP, and has set itself the target of increasing business expenditure on R&D to 1.67% by 2014 The recent introduction of tax credits for R&D should encourage greater participation in R&D amongst firms.20

2.33 Overall, the UK’s commitment to R&D is (or appears to be) mediocre and in aggregate

significantly below that of countries like the US, Japan, Germany, Sweden and Korea as depicted in figure 2.3 below (see Golborne, 2005 for a detailed analysis of R&D activities

by UK firms) Having said that, there is some reason to believe there may be greater under-recording of R&D in the UK than elsewhere, not least because before the introduction of tax credits there was little incentive for private firms to record their R&D

activities, and many appear unsure as to whether or not they undertake R&D (Howells et

al., 2001)

19 DTI (2003b) points out that the growth sectors within manufacturing have tended to be those that invest more

in R&D and have higher qualified workforces, as measured by the proportion of employees with degrees

20 As a whole the European Union has set itself the target of increasing R&D to 3% of GDP with two thirds of this expenditure coming from private sector enterprises (i.e business expenditure on R&D reaching 2%)

Figure 2.3 Gross Expenditure on R&D as a proportion of GDP, 2001 (OECD Data)

2.34 Patenting - Twelve countries (the US, Canada, Japan, Germany, the UK, France, Italy,

Switzerland, the Netherlands, Sweden, South Korea and Taiwan) account for about 95%

of all the US utility patents.21 Despite the strength of its science base, the UK’s share of

US utility patents has been in decline since the late 1960s, whilst newly emerging

countries like Korea and Taiwan are fast catching up with the UK in terms of patenting

(indeed, relatively to their size, these countries have arguably already surpassed the UK)

This situation is summarised in table 2.5 below

Table 2.5 Patenting in the United States – 1966 – 2000 – International Comparisons

US Utility Patents Granted Per 100 Patents Granted to the UK

2.35 The European Community Innovation Survey (CIS) – The CIS-3 found for the

period 1998 – 2000 that the UK had a lower than average proportion of enterprises

engaged in (technologically) innovative activities (i.e 36% c.f the EU average of around

40%) Only in Spain and Greece did lower proportions of enterprises report engaging in

(technologically) innovative activities A low UK propensity to engage in innovative

activities was found for all firm sizes, including large firms In the UK, only 57% of large

enterprises (those with 250+ employees) reported having engaged in innovation activities

21 Utility patents cover the useful features of an invention These patents cover a broad category of sciences Utility

patents have been issued for everything from anchors for buildings, zebra stripes as a method of camouflage,

computer software, to methods of doing business

between 1998 and 2000, compared with an EU average of 80% A lower propensity to innovate was also found for both industry (manufacturing and construction) and services

in the UK compared with the EU average. 22

2.36 According to the CIS-3 (Lucking, 2004), 21% of UK firms had introduced product

innovations compared to an EU average of 31% This deficit, relative to the EU average,

is again found to be consistent across both enterprise size and by sector of activity Meanwhile, amongst those enterprises that were innovation active, only 27% introduced a

“new to the market” product innovation, a proportion which is amongst the lowest in the Europe (EU average = 36%).23 Furthermore, just 17% of UK firms reported having introduced a process innovation between 1998 and 2000 This was the lowest proportion

of any country in Europe (with the EU average being 23%) UK firms’ engagement in process innovation was consistently lower than the EU average across all firm sizes and sectors of activity

Conclusions

2.37 This chapter has sought to establish the context for the review of the literature on the

relationship between skills and innovation which follows We have seen that the government has high ambitions for the UK: “We want the UK to be a key knowledge hub in the global economy, with a reputation not only for outstanding scientific and technological discovery, but also to be a world leader in turning that knowledge into new and exciting products and services In terms of business R&D and patenting we will aim

to be the leading major country in Europe within ten years” (DTI, 2003b, p 12)

2.38 Ultimately the relationships between skills, innovation and productivity, now and in the

future, are complex, with decisions made today impacting on performance in years and decades to come We can see this, for example, in the significant proportion of working age men without qualifications who are now economically inactive The Government’s challenge is to increase the productivity performance of the UK economy, whilst also ensuring that inequality does not widen further This involves both establishing the framework conditions within which business operate and, whilst holding down the overall tax burden, finding the best balance between various public investments For example, are further investments in the science base likely to be more effective (for efficiency and equality) than further investments in pre-school education? Ultimately such questions are probably impossible to answer with certainty, but it is important to recognise that the economy is such that investments in skills and innovation are often complementary, especially in the early stages of technological diffusion (see Chapter 3)

It is only by investing in both that large productivity gains are realised The balance is however often elusive In the remainder of this report we will explore the literature on the relationship between innovation and skills (in the context of raising value added and productivity) We will see that whilst much is known, there are also significant gaps in understanding

22 Averages here exclude the UK and Ireland

23 However, amongst those firms that did introduce product innovations, UK firms tended to claim the highest share of turnover was due to new products – at an average of 41% of total turnover This perhaps suggests a polarisation of innovation performance amongst UK firms, with some being highly innovative whist others have little or no commitment to product innovation

3 Innovation and Its Implications for the Demand for Skills

Introduction

3.1 In this Chapter we examine how innovation changes the demand for skills This Chapter

will focus primarily on contributions from ‘mainstream’ economics It should first be recognised that the effects of innovation on skills are among the most contentious issues

in economic theory In mainstream accounts of the problem, innovation has long been

narrowly defined as changes in technology which are taken as given in the economic system

and do not need explaining for the purpose of understanding their effects.24 This has favoured a strong emphasis on the demand side of the relationship between the supply of technical change and the demand for skills As a consequence, the main questions have been framed in terms of the impact of innovation on jobs, whereby firms decide on the relative advantage of acquiring more capital (i.e machinery, etc.) or more workers, and this balance can change as a result of the innovation.25

3.2 As Pianta (2005) recalls, the classic question is whether technology creates or destroys

jobs While this leads to analyses of innovation and employment levels, it also leads to investigations of the kind of jobs which are created and destroyed by innovation (i.e both

quantitative and qualitative questions are addressed) More specifically, questions on the

quality of employment concern how innovation or technological change affect:

– the composition of skills

– the distribution of wages

– the way these change over time within and across economies

Innovation and Employment

3.3 A traditional account of the microeconomic impacts of innovation on the quantity of

labour demanded is as follow In the search for higher profits, firms seek to reduce their intensity of use of the least productive factor of production (capital, labour etc.) in favour

of the most productive In the first place, they can do so by altering their technological endowments In a very simple framework, they can introduce innovations that make more intensive use of capital (i.e machinery, etc.) and reduce the proportion of human labour used, or they can introduce innovations that raise the (marginal) productivity of labour and diminish the incentives to invest in capital.26 This perspective on the

dynamics of investment in capital and labour is referred to as the theory of induced technical

change and has occupied economists since the pioneering formulations of Hicks

(1932/1963), Harrod (1939) and Samuelson (1965). 27

25 For a review and discussion of theoretical approaches, see Petit (1995) and Pianta (2005)

26 Of course, this is only half of the story because innovation always has the potential to change the nature of work

in such a way that existing workers can no longer perform it The questions this raises (for example, can these workers be retrained and at what costs?) will be dealt with in Chapter 4

27 Other related and by now ‘classic’ studies are Kennedy’s (1964), Drandakis and Phelps’ (1966), Griliches (1969), Fellner’s (1971) and Binswanger’s (1978), and Salter’s (1960) famous critique

3.4 Here ‘innovation’ is synonymous with process innovation and the broad picture is one where

if change takes the form of new machinery that reshapes the process of production, innovation tends to be associated with job losses The destruction of jobs can be compensated 1) by growth in the domestic and foreign demand for those goods and services they contribute to produce (in which case the aggregate number of jobs might increase, despite the fall in the proportional input of labour relative to machines) or 2) by the creation of new jobs either in the upstream production of new machines, in ancillary services, or in the unrelated production of new goods and services in other segments of the economy.28

3.5 This suggests that process innovation is conducive to technological unemployment in the

face of slow demand growth, but also that product innovation tends to be associated with

the creation of jobs Consequently a race is configured between the accumulation of capital which is reinvested by entrepreneurs in radically new businesses, which through product innovation tend to create new jobs, and the incentives to mechanise existing outputs through process innovation which tend to destroys jobs In theory, the overall effect of innovation on the demand for labour will depend on the balance between these processes

3.6 This has been explored empirically by Smolny (1998) who gathered evidence from a panel

of German firms and found that product innovation was associated with increased

employment while process innovation had no effect Meanwhile Brouwer et al (1993)

found, using a panel of Dutch firms, an overall negative relationship between innovation and employment but also noticed that more jobs were created where product innovations were more prominent And, analysing data on French firms, Greenan and Guellec (2000) found that both product and process innovations were associated with the creation of jobs at the level of the firm, but that the positive effects of process innovation disappeared at the industry level At the level of the firm the impact of product and process innovation on employment is not clear cut, and Pianta (2005, p 576) considers that is because “Firm level studies on the innovation employment link are unable to point out whether the output and job gains of innovating firms are achieved at the expense of competitors, or whether there is a net effect on aggregate industry employment It is often difficult to generalize beyond the groups of firms investigated and to compare results across countries When panel-data sets are used, a large part of the jobs created or lost may be accounted for by the entry or exit of firms left outside the panel.”

3.7 Industry-level studies (Meyer-Kramer, 1992; Antonucci and Pianta, 2002) generally

confirm the association between technological unemployment and process innovation, and employment growth and product innovation in markets where demand is expanding Pianta (2005), for example, shows the correlation between innovation and employment in France, Italy, the Netherlands and the UK for 20 industries over the period 1994-2000 (see Figure 3.1) Pianta shows that 1) overall, economies experienced job losses in the sectors analysed but with great variability across the sectors 2) high technology sectors with high rates of product innovation tend to cluster towards the top right quadrant, reflecting employment growth (the best performing sector being Information and Communication Technologies) and 3) more traditional sectors (e.g textiles, leather and footwear), which are usually characterised by intense process innovation and relatively high product substitutability, suffer more job losses and therefore cluster in the bottom left quadrant

28 This argument is as old as David Ricardo’s Principles of Political Economy and Taxation (1817/1921)

Figure 22.1 Share of new products in sales and employment change

Share of new products in sales, 1994-1996

.3 2

.1 0.0

UK14 UK13 UK12

UK10

UK09

UK08 UK07 UK06

UK05

UK04 UK03 UK02

UK01 NL20

NL16 NL15

NL14 NL13

NL12

NL11 NL10

NL09

NL08 NL07 NL06NL05

NL04 NL03

NL01

IT20 IT19 IT18

IT17 IT16

IT15

IT14

IT13 IT12

IT11 IT10

IT09

IT08 IT07

FR17

FR16

FR15 FR14

FR04

FR03 FR02 FR01

4%

Legend: Industry codes: 01 Food products; 02 Textiles; 03 Apparel; 04 Leather and footwear;

05 Wood; 06 Pulp and paper; 07 Printing and publishing; 08 Petroleum products; 09 Chemicals;

10 Rubber and plastics; 11 Non-metallic minerals; 12 Basic metals; 13 Fabricated metal products;

14 Machinery and equipment; 15 Office, computing machinery; 16 Electrical machinery;

17 Communications equipment; 18 Precision instruments; 19 Motor vehicles; 20 Other transport equipment Country codes: FR France; IT Italy; NL Netherlands; UK United Kingdom

3.8 To some extent, industry life cycles underpin the distribution of sectors depicted in the

chart In phases of early industry expansion, the rate of product innovation is high and the number of firms, employment and output tend to increase rapidly in spite of turbulent firm entry and exit patterns and highly heterogeneous product characteristics

As the industry matures, standards emerge which are often proprietary and scale economies grow in importance together with the size of the market So does employment, which also becomes more concentrated in fewer big firms, especially when the focus of innovation shifts from products to processes in order to cut costs in the face

of intense price competition For firms, this usually eventually results in the search for higher productivity through automation, the adoption of new technologies and a decreased reliance on large proportions of labour input We will discuss industry life cycles and patterns of innovation at greater length in Chapter 5

Innovation and Skill Biases

3.9 The effects of innovation on the type and quality of employment are no less important

than those on the quantity of employment, and have lately generated a growing interest among labour economists

3.10 Ideally, firms not only have to choose an optimal allocation of resources between capital

and labour but must also decide what kind of labour to use in combination with given capital (i.e machinery, etc.) endowments Their choices will depend on the relative scarcity, and hence the relative cost, of different kinds of skills, but also the productivity

of these different kinds of labour in relation to the kind of capital that is utilised in the production process If the technologies adopted by firms require higher proportions of

skilled workers over unskilled workers, technical change is said to be (positively)

‘skill-biased’.29

3.11 Figure 3.2 (from Machin, 2001) provides a simple illustration of the expected effects of

skilled-biased innovation on wage and employment rates of skilled and unskilled labour, assuming - for simplicity - that output is fixed, technical change is exogenous30 and supply conditions are stable The demand and supply curves are here constructed with the ratios between the wage of the skilled and the wage of the unskilled (vertical axis) and the ratio between the employment levels of the skilled and that of the unskilled (horizontal axis) The ratios represent, respectively, the relative wages and the relative utilisation of skilled

Source: Machin, 2001

3.12 If innovation is (positively) skill-biased, the relative demand for skilled and unskilled

labour will shift from D0 to D1 and lead to new wage and employment levels that augment the share of skilled workers to the detriment of unskilled workers With the limitations of all theoretical constructs, the shift is used in the literature to mirror the process of change observed in the last two or three decades in the labour markets of several advanced economies, where the demand for skilled workers appears to have

moved faster than the demand for unskilled workers, and faster than the supply of skilled

workers The reasons why the supply of skilled labour has been unable to keep pace with the rising demand for skilled labour will be discussed in the next chapter

3.13 The effects of technology on the demand for skills have been estimated in a large number

of studies over the last decade, as we shall see The phenomenon of skill-biased technical change has proven especially strong in the US, and to a lesser extent in the UK, where the number of skilled workers has been increasing fast and wage inequalities have grown wider across groups of workers with different skill endowments Evidence from

29 Acemoglu (2002) provides a clear and technically rich survey of the modern micro-economics of skill-biased technical change

30 That is coming from outside and not directly incorporated into or explained by the model

continental Europe is less clear-cut and the reason, it is generally argued, resides in the stricter regulation of labour markets on the continent, which mean wages do not smoothly adjust to demand forces with the consequence that technical change does not affect wage dispersion but translates instead in higher unemployment.31

How Technical Change Affects the Demand for Skills

3.14 The nature and direction of technology-related biases vary over time and it should be

emphasised that technical change is not inevitably (positively) skill-biased From a

historical viewpoint, as noted in Goldin and Katz (1998) and Manning (2004), at the beginning of the Industrial Revolution technical progress favoured the employment of unskilled workers and caused the demand for the skills of artisans to fall sharply during the 19th century Moreover, not only is technical change uneven over time but also across sectors of the economy (Malerba, 2005) As Manning (2004) remarks, there is a chance that, even after accounting for the effects of trade on employment, different and opposite changes in employment may be induced by different technological changes occurring at the same time in different industries Such effects would elude analysis at the macroeconomic level For example, there may be niches of the service sector (such as domestic cleaning and waiting tables) where technology favours unskilled labour due to 1) difficulties in mechanising complex hand-eye co-ordination tasks; 2) the impossibility of trade and 3) the growing demand of increasingly wealthy customers at the other end of the earnings spectrum for personal and customer services

3.15 Nevertheless, the opinion has emerged among labour economists that over the last

decades advanced capitalist systems have faced the emergence of a strong skill bias

associated with growing investment in capital that has led to increasing automation of

production tasks On the one hand, new machines have increasingly substituted for the labour of unskilled workers On the other, they have been introduced in combination with

the labour of highly skilled workers through which they can fully realise their productive potential In the following sections, we shall use a few selected examples to discuss the process of co-evolution of skills and innovation, first in the automobile industry, then in relation to the rise of information [and communication] technologies

The Paradigm of the Automobile Industry

3.16 Car production is a mature industry that well illustrates long-term processes of change in

scale intensive manufacturing Car manufacturing was born in the late 19th century as a craft-based business that required highly skilled and adaptable workers In the early 1900s, it was radically transformed by the diffusion of Taylorist methods of production, where complex tasks were simplified in such a way that unskilled workers could be made

to perform them after very little training The non-manual component of virtually all tasks was separated from the shop floor and concentrated in layers of specialised management (Kaplinski, 1988) The dramatic deskilling of manual labour was associated with the emergence of ‘Fordist’ mass production, the standardisation of components and increasing specialisation The growth in aggregate demand fuelled employment and absorbed increasing shares of unskilled labour.32

31 Piva et al (2005) provide a good overview of the recent literature on the topic and a discussion of international

comparisons We shall come back to this point in the section on Innovation, Skills and Wage Inequality

32 Note also that the cost of failing to educate the labour force to its full potential was not fully apparent, as during this Fordist era of mass-production there was a high demand for people without qualifications who could be persuaded to undertake highly repetitive forms of work (Prais, 1995)

3.17 If the introduction of the assembly line caused a relative shift from skilled to unskilled

labour, the emergence of Japanese ‘lean production’ techniques in the late 1970s started

to turn car production back into a fairly skill-intensive industry (Wells and Rawlinson, 1994).33 In the Japanese model, the introduction of multi-skill and multi-task jobs in mass-market industries reversed the global trend, led by US firms, towards excessive labour specialisation and favoured a distribution of skills across organisational levels Workers were encouraged to contribute to technical improvements This not only accelerated the reaction time to changes in demand, but also improved the innovative potential of firms by valuing workers’ creative efforts and by multiplying the opportunities for learning and for the incremental acquisition of complementary skills (Kaplinski, 1988) See Case Study 1 below on Japanese Production Practices in the Car Industry: Implications for Skills

Various studies have shown that Japanese production systems tend to be highly collaborative

with an emphasis on team working According to Clark et al (1987) in the Japanese system: “a

heavyweight project manager leads a multifunctional team, in which problem-solving cycles are overlapping and closely integrated through intensive dialogue” (p 766) The heavyweight teams had a small number of multifunction engineers which facilitate effective worker communication and so the efficiency and quality of design In contrast, some of the US firms had overspecialised their engineers – one firm had specialist door lock designers – and drove up the size of teams, diminishing communications within the team (Howells, 2005, p 188) Small, highly co-ordinated teams, permit the shift from sequential developments to parallel development (or ‘simultaneous engineering’) processes.34 Also in contrast to Western firms, Japanese graduate engineers typically begin their careers working on the shop floor Consequently, the shop floor became something of a laboratory employing a high concentration

of engineers as technology implementation problem solvers Engineers then tend to rotate around the functional departments of the firm, which encourages an emphasis on practical skills, broadly applied, rather than abstract knowledge applied in specialist niches “The achievement

of the necessarily small, heavyweight teams in design was possible because the members already shared much experience of each other’s nominal functional specialities This was an obvious aid

to communication, but it also allowed fewer individuals to represent the necessary fields of expertise” (Howells, 2005, p 189) To summarise, this organisational innovation resulted in an increased need for teamworking, communication, and broadly applied practical skills in problem solving amongst the engineers, at the expense of narrow specialisations

Also crucial in the Japanese system is the link, provided by production supervisors, between management and the workforce Traditionally in the UK supervisors had tended to be promoted from the shop floor on the basis of on-the-job performance, but had few formal qualifications Their generally poor performance as supervisors had led to their being progressively stripped of responsibilities, which were reassigned to professional specialists: so personnel specialists dealt

33 See Womack et al (1990) for the thorough analysis of the evolution of the sector and Utterback (1996) for a

discussion of it from an innovation perspective

34 “The key question is how the Japanese kept their design teams small and effective In the first instance, small size was only possible because of the breadth of functional knowledge possessed by the individual engineers in the team: fewer individuals were needed to cover the necessary areas of expertise This only begs the question how this was achieved” (Howells, 2005, p 189)

with industrial relations; specialist layout and maintenance engineers had control over those functions; shop stewards controlled work organisation and the division of labour Over time, this reduction of responsibilities led to the erosion of supervisors’ pay differential over shop floor workers, a decline in the respect given to supervisors, which made it increasingly difficult to assign capable people to supervisory roles When it started its new plant near Sunderland, Nissan first doubled the pay of supervisors to the level of a professional engineer It then widened the supervisor’s responsibilities to include: quality control; hiring of workers; maintenance; work area layout; design of the cost control system and work scheduling, so returning all the basic operations and responsibilities to this pivotal managerial position To recruit supervisors, Nissan used sophisticated psychological and personality tests to select people who could creatively discharge their responsibilities Of the 22 new supervisors Nissan recruited, only six had management experience in the British car industry As with the engineers, this organisational innovation had the result of increasing the need for a broad spectrum of skills at first at the line manager level, at the expense of demand for specialist skills elsewhere in the firm (Wickens 1987; Howells, 2005)

3.17 Consequently, during the 1980s in the UK, jobs that used to be performed by the

unskilled became increasingly automated and tasks that were routine became complicated

by the adoption of robotics and therefore started to require a higher percentage of higher skilled staff Recruiting and training practices changed accordingly From the quick, task-specific and on-the-job methods of the Fordist period, a shift took place towards formal education and lengthy ‘before-the-job’ search and selection processes aimed at testing thinking (cognitive) and diagnostic skills of prospective employees (Gilles, 1978, and Rutherford, 1994; cited in Duranton, 2003) Furthermore, an increasing number of white collar workers were employed in managerial positions and highly qualified workers were increasingly employed in applied research and product development activities for the production of increasingly technologically complex products Also, a shift has been observed from the production of goods to the provision of complementary services (Howells, 2003) For car producers, this has implied diversification of their business portfolios to include financial and leasing services, activities that – until now – have employed mainly skilled graduates (in Chapter 5 we refer to this as the “creative accumulation mode of innovation”)

3.18 Thus, to take a general perspective over the long run in advanced economies such as the

UK, the mechanisation of manufacturing activities in the automobile industry has dramatically reduced the demand for unskilled workers with adverse consequences for their wages and employment rates (Duranton, 2003; McLoughlin and Clark, 1994) At the same time, innovation has led to the rising employment of higher-skilled workers that are able to use, maintain and improve the new technologies and organisational models the innovation involves This equates to a rising demand for high level skills Finally, the impact of trade should not be underestimated The effects of technical change are in fact deeply connected to those derived from increasing global competition, which has dramatically augmented pressures on costs and performance for incumbents in mature markets and relatively high cost places of production, such as the UK More specifically, newly industrialising countries - and China in particular - are absorbing increasing shares

of the global market for the production of finite goods as well as the low-cost production

of standard components that are later used in assembly elsewhere This has further accelerated the shift in the composition of labour and strengthened the incentives for western manufacturers to pursue high value-added strategies These typically involve strong emphases on research, design-intensive product development, high-technology content of modular components, and heavily skill-intensive innovation strategies

3.19 It is then quite clear that processes of substitution of unskilled labour with machinery are

only part of the story of skill-biased technical change The simple observation that modern production practices are associated with increasing employability and increasing wages amongst the high skilled suggests that capital investments and investments in skills are complementary (especially with respect to cognitive and relational tasks.)

The Rise of Information Technologies

3.20 Nelson and Phelps (1966) and Griliches (1969) suggested that adaptable skills are

themselves essential in the process of adoption of new technologies and that the diffusion of innovations will tend to favour economies, firms and individuals with higher human capital (see Chapter 4) This is particularly the case with the diffusion of advanced technologies, for which there is a greater need for high absorptive capacity (Cohen and Levinthal, 1990), that is the capacity to learn Berman, Bound and Griliches (1994) and Levy and Murnane (1996) tested this hypothesis and found positive correlations between intensive use of skills and investments in capital-embodied new technologies (machinery) across industries and across plants within industries Also at the plant level, Doms, Dunne and Troske (1997) found that the number of new technologies in use is associated with employment of more skilled workers and higher average wages However, longitudinal analysis of the data did not provide evidence that the adoption of new technologies is significantly correlated with skill-upgrading, although this was the case with investments in information technologies

3.21 As Bresnahan et al (2002: 340) argue, “…the size, breadth and timing of the recent labour

demand shift have led many [researchers] to seek skill-biased technical change in the largest and most widespread technical change of the current era, information technology” Indeed, the majority of studies on the skill-bias hypothesis focus on the adoption of IT equipment at different levels of aggregation: at the individual level (for

example, Krueger; 1993), at the firm level (Doms et al., 1997; Mairesse et al., 2001) and at the industry level (Berman et al 1994; Autor et al 1998; Falk, 1999) The vast majority of

contributions in this area talk about complementarities between capital investment and skills to explain the effects of the diffusion of IT on wage differentials and employment patterns within and across sectors These effects are especially clear for the 1980s, when the greatest skill-upgrading was recorded in those manufacturing industries where the introduction and diffusion of microprocessor technology was more rapid (Berman, Bound and Machin, 1998) However, greater uncertainty exists for studies relating to the 1990s, which perhaps indicates that IT investments have moved to a more mature stage, and thus have a lower requirement for skills

3.22 The core of the argument connecting investment in IT and skills (or human capital) is

that better-educated workers are able to extract better performances from the use of computers than less educated ones This was especially true in the early phase of the diffusion of computers, before the introduction of graphical inter-faces and ‘point-and-click’ technologies, which made computers a lot more accessible to those without specialist skills As a consequence, better educated workers receive wage-payoffs that are proportional to their increasing marginal productivity in a labour market where the demand for computing skills is yet unsaturated and perhaps accelerating A number of empirical studies on the impact of the IT revolution have found strong correlations over time between IT investments, productivity and higher wages to skilled workers In their investigation of changes in the structure of wages in the US during the 1980s, Bound and Johnson (1992) reasoned that the rising wage inequality could not be attributed to the increase in the trade deficits, to the declining power of unions or to a fall in the growth

rate of the supply of college-graduates As a consequence, they argued that explanations had to be found in changes in technology and in particular in the diffusion of computers 3.23 Juhn, Murphy and Pierce (1993) also measured the increase in wage inequality, which

reflects substantial shifts in the demand for labour, among US male workers between

1963 and 1989 and related this to the returns on observed (i.e education and experience) and unobserved factors They found that most of the variation was to be attributed to unobserved components and felt this was likely to consist of technical progress and changes in the “world economy”

3.24 Krueger (1993) reached similar conclusions in a classic analysis of detailed microdata for

the period 1984-1989 He decomposed the increase in the rate of return to education by using direct measures of the use of computers and estimated that between one third and one half of the increase was related to computer usage This also accounted for an observed 10-15 percent rise in the wage rate of computer users In another very influential study, Autor, Katz and Krueger (1999) conjectured that better-educated workers could achieve bigger productivity gains with the continuous improvement of computer technology They also discussed the changes in relative supply of and demand for skills in the US from 1940 to 1996 and argued that for the period 1960-1990 computer-related technical change was the major determinant of the observed shift in the demand for skills

3.25 However influential, these conclusions are not uncontroversial and have occasionally

been commented upon with scepticism Di Nardo and Pischke (1997) question the supposed causal relationship between computer usage and wage differentials Technically,

it is possible that wage differentials are not due to IT-related variables but reflect instead unobserved heterogeneity among workers While high correlations can indeed be found, their interpretation can go either way and computers may be associated with workers with higher skills because these are better paid and, as a consequence, allowed access to better equipment independently of specific productivity gains Krueger had already considered this possibility in his 1993 study, but expressed a preference for the skill-bias hypothesis

To test their hypothesis, Di Nardo and Pischke (1997) apply to German data the same econometric tools of Kruger’s to estimate the effects produced on wages by computer usage and those eventually produced by the use of telephones, pens or pencils, or from sitting on the job They find strong and significant effects and conjecture that as it is

unlikely that such gains are obtained through the use of pencils or by sitting, the claim

that these would be due to computers is equally unwarranted Dickerson and Green (2004) discuss in some detail the problem of estimating the value of computing skills and use UK Skills Survey Data for 1997 and 2001 to assess the effect of the use of computers

on wages Their results show that when controlling for a broad range of job attributes and the complexity of the tasks involved, computer usage, together with the complexity

of the way computers are used, remains positively and significantly associated with higher pay while the effects of tasks such as reading or writing short documents on the job become less significant

3.26 A balanced assessment of the findings should probably consider that while strong

correlations are found between IT and skills, the identification and estimation of causal links is not without problems Second, it is important to emphasise that as a technology matures, the demand for skills is likely to change For example, before the introduction

of graphical interfaces, ‘point-and-click’ technologies, and standardised menu driven computer software, the use of computers usually required at least some knowledge of one

or more computer languages.35 Third, the transformation of business brought about by the IT revolution is probably better understood in connection with a broader view of changes that have occurred in the organisation of labour and in the structure of production processes, to which we now turn

How Organisational Innovation Affects the Demand for Skills

3.27 Traditional definitions of innovation concerned only new products and processes and

carried strong technological overtones However, ‘softer’ forms of innovation are increasingly recognised as fundamental to the growth and sustainability of profits The way in which the system of production and provision of goods and services is structured

is pivotal to its competitiveness At the level of the firm, this means that the organisation

of the tasks involved in running a business is a precondition for its success In dynamic terms, it also means that the chances of the firm surviving in competitive environments depend not only on its capacity to innovate products and processes, but also on its capacity to transform its established managerial practices and its internal structure Yet it

is now clear that the study of organisational innovation has lagged far behind research on product and process innovation (Lam, 2005) although increasing attention is being paid

to the effects of skill biased organisational change, which is found to complement the effects of technical change on employment, job-quality and wages

3.28 Again with respect to IT, Bresnahan (1999) articulates the problem of skills and

innovation in the following way He acknowledges the weakness of the most popular estimation procedures for evaluating the impact on productivity of IT, but argues that this is not enough to refute the hypothesis of computer-based skill-biased technical change He suggests that IT has raised the demand for skilled workers but has done so

by changing the organisation of the firm and not so much by improving the productivity

of individual workers Bresnahan dismisses the idea that computers and human capital are direct complements and stresses instead that complementarity arises in a systemic way

only at the firm level He therefore speaks of limited substitutability and organisational

complementarity between IT and skills The first point concerns the imperfect capacity of

computers to substitute non-routine ‘cognitive’ and ‘interactive’ features of labour (human decision-making above all); the second concerns the way in which the application

in workplaces of computers - and especially networked computers - changes production

processes and often the form of production processes (as in many service industries).36 This implies that understanding trends in the demand for skills requires consideration of

the links between organisational change and the changing nature of work

3.29 It also means that the internal and external structure of the firm deserves more attention

because it is a strong determinant of firm performance The external organisation of the firm (Richardson, 1972) consists of its connections with other organisations operating in the relevant innovation systems Outward connections include supply relationships, business partnerships, collaborations with universities and contact with consumers and regulatory agencies The internal organisation of the firm consists, instead, of boundaries

of competences, as well as the structure of connections between them, which reflect the division of labour and the allocation of tasks between teams, divisions or plants The emergence of IT has clearly opened up tremendous opportunities for restructuring the

35 This pattern fits with the classic life cycle where skill requirements diminish with the maturity of the technology

36 As a consequence, estimation of individual wage equations with a regressor for computer use is not an accurate tool to measure the impact of computers on the labour market

channels through which information circulates from, to and within the firm, and for reshaping the division of information-based labour and the way this is carried out

3.30 Greenan and Guellec (1998) found a strong positive correlation between organisational

change and up-skilling at the level of the firm in French cross-sectional survey data for the late 1980s Caroli and van Reenen (2001) find the same results using panel data on British and French firms for several years between 1984 and 1992 Greenan (2003) also uses a survey conducted among French manufacturing firms in 1993 to test the correlations among technical change, organizational change and skill change.37 She finds,

on the one hand, that both new manufacturing technologies and organizational changes can destroy jobs; the former by replacing labour with capital, the latter by rationalizing the use of labour On the other hand, however, the combination of the two can boost employment when it is aimed at seeking market expansion Greenan combines data on the acquisition of new technologies for six advanced manufacturing processes and finds positive correlations across all of them, the strongest impact being between computer-aided production systems and stock control The analysis of organizational change reveals that a significant class of firms moved towards increased flexibility Finally, the distribution of skills seems to have been biased towards increased expertise and managerial control of centralized firms Therefore, in general, the results confirm a strong complementarity between technological and organizational change

3.31 In the same analysis (Greenan, 2003) skill changes emerge as more closely connected to

organizational than technical change In turn, the balance between the organizational and the technical dimensions has relevant consequences for the demand for skills Firms adopting advanced manufacturing technologies without changing their organization will just increase the employed labour force Meanwhile, the firms that underwent reorganization experienced an evolution in the occupational structure as well as in the content of the job The categories that are observed to gain responsibilities are blue collar workers in the switch towards the flexible-enterprise model, specialists when the adoption of new technologies involves expertise, and management when the hierarchical structure is changed These differences, however, are not reflected in employment levels 3.32 Bresnahan et al (2002) also report a strong correlation between (positive) skill-biased

technical change and organisational change with the introduction of IT equipment in a sample of US firms They argue that “…firms do not simply plug in computers or telecommunications equipment and achieve service quality or efficiency gains Instead they go through a process of organizational redesign and make substantial changes to their product and service mix […] … IT is embedded in a cluster of related innovations, notably organizational changes and product innovations These three complementary innovations – 1) increased use of IT, 2) changes in organization practices, and 3) changes

in products and services – taken together [emphasis in the original] are the skill-biased technical change that calls for a higher-skilled labour mix.” (Bresnahan et al., 2002, pp 340-341) Piva et al (2005) have recently reached the same conclusion on a sample of

Italian firms where up-skilling went hand in hand with both technical and organisational change

37 Precise definitions are given in this study for these three dimensions Technical change is referred to changes in the capital factor such as the replacement of old equipment by new one Organizational change involves changes in the labour factor such as investments in new skills and competences Skill changes are related to the use and

combination of these factors, thus redefining the division of labour, the content of the job within newly introduced processes and the underpinning rules of coordination

3.33 This seems especially relevant for the service sector, where innovation is still relatively

understudied, as stressed in Miles (2005) It can be argued that where the boundaries between process and product innovation are rather ill defined – as is often the case with services – complementarities between technical and organisational change acquire paramount importance in shaping the distribution of skills within and across firms The growth in outsourcing, for example, means that activities that were previously undertaken within firms are now being undertaken by others, or in collaborative arrangements between firms This requires that employees have to learn new ways of working and how

to agree contracts An overly rigid contractual situation will often stifle the relationship and reduce the effectiveness of the partnership See Case Study 2 on Innovation and Co-operation in UK Transport Services below

The UK has some of the most congested transport systems in the world, not least in its airport and railway infrastructures, the utilisation of which has grown well beyond the design limits envisaged by even the most optimistic planners In both cases, and especially around London, this reflects an unwillingness to build new infrastructure to meet growing demand For example,

in contrast to Paris Charles de Gaulle which has expanded over the years and now has four parallel runways, Heathrow has just two parallel runways, the same basic infrastructure it has had since the 1950s

In the absence of new infrastructure, the BAA, in conjunction with National Air Traffic Services and the airlines have explored new means of expanding the capacity of Heathrow and the UK’s other congested airports This has involved the development of co-operative practices, and new skills in co-operating and ‘heedful inter-relating’ amongst air traffic controllers and airline pilots, and has led to year on year improvements in the utilisation of Heathrow’s runways, despite the increased use of large aircraft and the need to maintain the existing quality of service (Tether and Metcalfe, 2003).38 Air traffic controllers have had to learn new skills for aircraft sequencing and traffic management, whilst pilots have had to learn to expedite their operations within the airport system to maximise its total utilisation rather than with a view to their individual performance

In the railways, Network Rail appears to have developed a similar co-operative approach with the Train Operating Companies (TOCs) which is paying dividends in terms of increased punctuality and reduced delays This is a marked contrast to the days of Railtrack, which attempted to manage the system through highly specified contracts

The new climate of cooperation has been accompanied by the setting up of joint improvement teams - such as the one that has helped Midland Mainline improve punctuality by 20 per cent in

a year - and joint control centres, where TOC and Network Rail managers sit side by side to oversee operations In time the hope is that growing trust will lead to the establishment of truly integrated teams under one overall manager.39 This implies the development of co-operative and negotiation skills, rather than just ‘command and control’ skills

3.34 The importance of the combined effects of technical and organisational change is then

especially clear for the service sector, which tends to make intensive use of technology (Pavitt, 1984) However, evidence appears to be mixed on the rate and direction of biases

in the composition of the demand for skills While highly educated workers dominate in knowledge-intensive sectors, such as consultancy services, the same cannot be said of those parts of the service economy where the business relies on routine tasks that do not require special qualifications To some extent, these routine tasks may be vulnerable in the face of technical progress (i.e automation) and growing incentives to relocate in lower cost regions (for example, back-office work and call centres’ activities in India) However, the extent to which services can be affected by automation and off-shoring remains to be seen

Soft-Skills and Emotional Labour

3.35 Other forms of innovation involve developing skills which have little or nothing to do

with technology For example, adding value through services often involves developing empathetic relationships between the service provider and the customer, and this has led

to the recognition and growing importance of ‘emotional labour’,40 as a core component

of some forms of service work, such as airline stewards (Williams, 2003), tour reps and nurses (e.g McQueen, 2004) James (1992) for example, states: “The formula ‘care = organisation + physical labour + emotional labour’ identifies component parts of

‘carework’ as observed at a hospice New forms of service – such as image consultants (Wellington and Bryson, 2001) - have developed around ‘emotional labour’, and call centre staff are expected to act down the phone (Palmer and Carstairs, 2003) Meanwhile more established activities such as beauty therapists increasingly recognise themselves as engaging in emotional labour Sharma and Black (2001) found that beauty therapists saw their work less in terms of what it does to make women look better, more in terms of what it does to make women feel better It follows, therefore, that knowledge of, and the physical ability to undertake beauty therapies is necessary but not sufficient to survive in that occupation

3.36 Similarly, ‘emotional labour’ is increasingly recognised as necessary amongst

professionals It is not sufficient that barristers who contest jury trails be experts in the law – they also need to be able to emote to make an emotional connection with the jury (Harris, 2002) Being an expert in your field is no longer sufficient for university lecturers, for as student feedback becomes ever more important those lecturers who gain are those who provide dynamic presentations (in PowerPoint, etc.) and entertain, as well

as (perhaps) educate (Ogbonna and Harris, 2004)

3.37 In frontline service work, such as in banks, call centres and amongst airline stewards,

‘emotional labour’ extends beyond ‘good manners’, to include a complex of skills including communication skills, the ability to empathise and understand problems from another’s (usually the customer’s) perspective, the ability to calm irate customers, the ability to solve problems, the ability to work in teams with flexible job specifications, and the ability to reflect on performance (Gorman, 2000) The point is that workers are often expected to conceal or manage their own feelings for the benefit of a successful service delivery (Constanti and Gibbs, 2005)

3.38 As these services have expanded, the supply of this labour has become increasingly

constrained relative to demand Moreover, as some companies have developed more

40 ‘Emotional labour’ was first highlighted by Hochschild in ‘The Managed Heart: The Commercialisation of Human Feelings’ (1983)

defined ‘customer service’ strategies, their demands for ‘emotional labour’ have become more exacting or specific, particularly amongst those companies that rely on ‘full service strategies’ to compete against the reduced, ‘no-frills’ approach of new rivals Consequently, these skills have become more defined, and often customer specific, rather than ill-defined and generic

Complementary Explanations of the Skill Bias

3.39 Labour economists tend to agree that (positively) skill-biased technical change has

contributed to altering the distribution of earnings, to exacerbate wage inequalities and/or

to increase economy-wide or sector-specific unemployment Yet, this is not the only factor that is relevant in co-determining the quantity and quality of the demand for labour The most important complementary explanation is trade Globalisation is often used to account for the changes in occupational trends in advanced OECD countries, as documented for example in Nickell and Bell (1995) The effects of trade on the demand

for labour consist of sector-biases (as opposed to skill-biases) whereby increasing/declining

trade in different sectors impacts on the relative demand for capital and labour, since their relative demand varies by sector If this explanation dominates the evolution of the demand for labour, shifts of labour should be observed between, and not within, sectors (see Woods, 1994; Krugman, 2000; Heskel and Slaughter, 2002)

3.40 The essence of a pure trade-theoretic argument is as follows In the last 20 years, trade

flows between developed and developing countries have increased significantly Consider

a world with one developed country and one developing country, both with two sectors,

a skilled-labour intensive sector and an unskilled labour intensive sector The developing country abounds with unskilled labour of very low unit cost, whilst the developed country abounds in skilled labour The developing country therefore increasingly specialises in the production of unskilled-labour intensive goods, as trade is opened up and the developed country increasingly specialises in skilled-labour intensive goods The availability of unskilled-labour intensive goods in the developed country through trade diminishes the demand for the same goods produced at higher costs domestically With the contraction

of the demand for the developed countries’ unskilled-labour intensive goods, the demand for their domestic labour and capital is also reduced, and especially the demand for the least productive factor (which is unskilled labour) This results in the transfer of workers from the unskilled-labour intensive sector to the sector where the country maintains a comparative advantage, i.e the skilled-labour intensive sector The resulting oversupply

of unskilled labour drives down their wages As a consequence, in this simple country two-sector framework the rise of wage inequality is not due to biases in technical progress (i.e innovation) but instead to the differential export performances of sectors of the economy competing for productive resources

two-3.41 Machin (2001) summarises the main arguments against the view that trade explains

employment and wage patterns He remarks that, however plausible, this explanation alone would neglect the fact that most trade still takes place between developed countries and that the volume of transactions between developed and developing countries appears

to be too small to account for the widening inequalities observed in the labour markets of developed countries Second, it would not explain why inequalities are widening both in

developed and developing countries Third, it would fail to notice that industries where

trade has accelerated are not those where the share of skilled to unskilled workers has increased faster Also, in non-manufacturing industries where the outputs have been difficult to trade internationally –there have been similar rates of skill-upgrading A further point is that this explanation assumes that international trade is between sectors, so-called inter-industry trade, where a country imports in some sectors and exports in

others However, it has been observed empirically that most trade is now intra-industry trade, i.e where each country both imports and exports within the same industry

3.42 In their study of the demand for labour in US manufacturing, Berman, Bound and

Griliches (1994), found that the employment patterns (reflecting the demand for skills) changed more within, than between sectors They attributed the within sector changes to technological innovation, and the between sector changes could be attributed to trade; this evidence therefore suggests innovation is more important in shaping the demand for skills than trade.41 Katz and Murphy (1992) reached essentially the same conclusion (see also Johnson, 1997) Machin and Van Reenen (1998) also report little evidence of effects

at the industry level of import intensity on shifts in the proportion of skills in 7 OECD countries (including the US and the UK).42

3.43 Gregory et al (2001) estimate the sources of the changing structure of skills, and in

particular the extent of the skill shifts due to either trade or technical change, by using an input-output model for the UK economy after 1980.43 The study highlights the importance of inter-sectoral trade mechanisms in the determination of the final levels of output and employment By considering relative skill endowment a key determinant of trade, they explain the major dynamics observed in the shifting composition of the labour force in the UK Such a wide perspective also allows for important considerations on the growing role of business services as a major component of inter-industry trade The main findings are that technological change has the greatest role in explaining the shifts of the skill composition in the UK - that is, the increased demand for skilled work against falling demand for low-skilled workers The disaggregation at sectoral level is particularly important here as it provides a measure of the relative importance of services as intermediate inputs The results indicate that domestic consumption is one of the chief engines that have boosted skilled employment in the UK, while technological change is most responsible for the loss of low-skill jobs

3.44 Hoskins (2000; 2002) also looks into the relative importance of the causes of structural

changes in the composition of employment in the UK for the period 1951-1991 and finds strong differences between skilled and unskilled workers as well as shifts between skill categories within each group Hoskins shows that the effects of technology after the 1960s are negative among skilled non-manual workers but positive for higher ranked types of skilled occupations (especially managers and professionals) In addition, both sectoral (arguably driven by trade) and technological change are found to be important for all groups in all periods, but also that the balance between their relative importance varies with a significant shift towards skilled occupations

3.45 Although the debate is often framed in rather antagonistic terms between labour and

trade specialists, as Haskell (1999) and Nickell (2004) argue, technology and trade are broadly complementary explanations of shifts in the demand for labour Furthermore,

41 They also attributed part of the variation to the increase in defence procurements that characterised US policies

of difference between the last and the first year of observation