Technical ReportPDF Available

Abstract and Figures

This guidance paper provides a common intellectual understanding for the BEYOND4.0 research project. It explains what is to be analysed, why and how. As part of this task, it explains the key developments, issues and concepts that drive the project. It provides a common starting point for the aim of BEYOND4.0 to support the delivery of an inclusive European future by examining the impact of new technologies on the future of jobs, business models and welfare in the European Union (EU).
Content may be subject to copyright.
This project has received funding from the European Union’s Horizon 2020
research and innovation programme under grant agreement No 8222293.
D2.1 Guidance paper on key
concepts, issues and developments
Conceptual framework guide and
working paper
Version 1.0
Date: November 2019
Chris Warhurst, Sally-Anne Barnes & Sally Wright with Steven
Dhondt, Christine Erhel, Nathalie Greenan, Mathilde Guergoat-
Larivière, Sylvie Hamon-Cholet, Ekaterina Kalugina, Olli E. Kangas,
Vassil Kirov, Michael Kohlgrüber, Christopher Mathieu, Tasmin
Murray Leach, Peter Oeij, Carlota Perez, Egoitz Pomares, Joshua
Ryan-Collins, Antonius Schröder & Frans van der Zee.
File: BEY4.0_WP02_Task_2.1_guidance_paper_FINAL_20191112
Status: Approved by EB on 7 November 2019
Table of contents
EXECUTIVE SUMMARY ..................................................................................................................................... I
1. INTRODUCTION .......................................................................................................................................1
1.1. AIM OF BEYOND4.0 ................................................................................................................................. 1
1.2. OBJECTIVES OF BEYOND4.0 ....................................................................................................................... 1
1.3. PURPOSE AND STRUCTURE OF THE GUIDANCE PAPER ......................................................................................... 3
2. FOCUS OF BEYOND4.0 - INDUSTRIE 4.0 AND UBERISATION .....................................................................4
3. GENERATING THE EMPIRICAL DATA ........................................................................................................6
3.1. WORK STRANDS ......................................................................................................................................... 6
3.2. OUTCOMES ............................................................................................................................................... 8
4. THE AGE OF REVOLUTIONS? ....................................................................................................................9
4.1. TECHNOLOGICAL REVOLUTIONS ..................................................................................................................... 9
4.2. THE DIGITAL REVOLUTION ......................................................................................................................... 14
4.3. THE IMPACT ON WELFARE POLICY ................................................................................................................ 19
5. KEY CONCEPTS AND ISSUES INFORMING THE FOUR AREAS OF ENQUIRY .............................................. 24
BEYOND4.0. THE QUALITY, CONTENT AND DISTRIBUTION OF WORK .............................................................................. 24
5.2. THE SKILL NEEDS OF THE LABOUR MARKET ..................................................................................................... 31
5.3. EDUCATION AND TRAINING TO SUPPORT SKILL DEVELOPMENT ........................................................................... 36
5.4. THE CREATION AND CAPTURE OF VALUE BY COMPANIES .................................................................................... 40
6. CONCLUSION ......................................................................................................................................... 47
REFERENCES ................................................................................................................................................... 48
Executive Summary
This guidance paper provides a common intellectual understanding for the BEYOND4.0
research project. It explains what is to be analysed, why and how. As part of this task, it explains
the key developments, issues and concepts that drive the project. It provides a common
starting point for the aim of BEYOND4.0 to support the delivery of an inclusive European future
by examining the impact of new technologies on the future of jobs, business models and
welfare in the European Union (EU).
The main premise for BEYOND4.0 is that technology is not deterministic but socially negotiated
by key social actors at various levels: firms, industry, regional, national and EU. This premise
opens up the possiblity that the EU can use digital technologies, specifically those related to
Industrie 4.0 and Uberisation, to promote the creation of an inclusive digital economy that
provides decent work and decent lives for EU citizens.
To date, such integrative research at the EU-level is lacking. In meeting the need for such
integrative research, the objectives of BEYOND4.0 are fivefold:
1. Provide new and systematic scientific insight into technological transformation and the
extent to which digital transformation is axiomatically disruptive.
2. Provide new and systematic scientific insight into company strategies dealing with
technological transformation, examining the variety of strategies and their expected
impacts as well as the role of social dialogue with key actors at various levels.
3. Examine the impact of this technological transformation using new, innovative approaches
to analyse and predict its impact.
4. Identify the range of policy options to deal with the consequences of technological
5. Identify the range of social investment approaches and tools that support a form of desired
technological transformation that is inclusive.
The new empirical data generated by BEYOND4.0 has three outcomes. The first is new scientific
understanding of the impact of the new digital technologies in relation to work and welfare.
The second is diagnostic and development tools to lever technological opportunities. The third
is evidence-based support for social and competitive EU policy.
With these aims, objectives and outcomes, BEYOND4.0 better addresses the challenges of the
putative 4th Industrial Revolution and seeks to provide alternative policy options for the EU. In
doing so, BEYOND4.0 provides new insights and measures to help address poverty, equality
and decent work.
Bookended by an Introduction and Conclusion, the paper has four main sections. The first
outlines BEYOND4.0’s focus on Industrie 4.0 and Uberisation. The second explains the empirical
research design f the project. The third elaborates debates about technological revolutions,
including the current digital one, and the latter’s implications for work and welfare. The fourth
presents the key concepts and issues around the four areas of empirical research enquiry for
BEYOND4.0: the quality, content, and distribution of work; the skill needs of the labour market;
the education and training to support development of these skills; and the creation and
capture of value by companies.
1. Introduction
The BEYOND4.0 research project focuses on the evidence-based delivery of an inclusive European
future by examining the impact of new technologies on the future of jobs, business models and
welfare in the European Union (EU). In this respect, the project is deliberately ambitious in its scope
and intent. This guidance paper sets out the general starting point of the project in terms of the key
concepts, issues and developments driving the BEYOND4.0 project. It outlines BEYOND4.0’s aims
and objectives, the digitalisation developments that are the focus of the project and the four areas
of enquiry following these developments.
1.1. Aim of BEYOND4.0
The overarching aim of BEYOND4.0 is to help deliver an inclusive European future through an
evidence-based examination of the impact of the new ditigal technologies on the future of jobs,
business models and welfare in the EU. This examination centres on ‘Industrie 4.0’ and ‘Uberisation’
as emblematic of these new technologies.
The main premise for BEYOND4.0 is that technology is not deterministic but socially negotiated by
key social actors at various levels: firms, industry, regional, national and EU (Berting, 1993; Bijker et
al., 2012; Child, 1972; Noble, 1984). This premise then opens up the possiblity that the EU can use
these technologies to promote the creation of an inclusive digital economy that provides decent
work and decent lives for EU citizens. To date, such integrative research at the EU-level is lacking.
Indeed much of the current debate about the impact of the new digital technologies is hypothetical
and speculative, lacking empirical evidence. Moreover, in general, too much social and economic
science analyses shy away from displaying any normative interest in improving (working) lives
(Grote & Guest, 2017).
1.2. Objectives of BEYOND4.0
The objectives of BEYOND4.0 are fivefold:
1. Provide new and systematic scientific insight into technological transformation and the extent
to which digital transformation is axiomatically disruptive.
2. Provide new and systematic scientific insight into company strategies dealing with technological
transformation, examining the variety of strategies and their expected impacts as well as the
role of social dialogue with key actors at various levels.
3. Examine the impact of this technological transformation using new, innovative approaches to
analyse and take into account:
Quality (including occupational safety and health OSH), content, and distribution of
work, as well as the distribution of work amongst different types of workers (e.g. gender,
age, skill level, geographic location);
Skill needs as function of the digitisation of task content rather than occupational
activities; Skill utilisation as a function of organisational choices that favour or impede
continuous on the job learning and re-arrangements of skill sets;
Education and training to support skill development, so sustained employability and
opportunity for social mobility are enhanced;
The creation and capture (extraction) of value (and wealth) by companies.
4. Identify the range of policy options to deal with the consequences of technological
transformation, in terms of:
Fiscal policy (e.g. the introduction of taxes on new forms of rent extraction enabled by
robot taxes);
Welfare policy (e.g. experimentation with basic income and other innovative social
security approaches).
5. Identify the range of social investment approaches and tools that support a form of desired
technological transformation; that is, policy centred on a ‘high road, inclusive approach to
Figure 1 provides a graphical illustration of the five inter-connected objectives.
Figure 1: BEYOND4.0 objectives
With these aim and objectives, BEYOND4.0 better addresses the challenges of the putative 4th
Industrial Revolution and seeks to provide alternative policy options for the EU. In doing so,
BEYOND4.0 provides new insights and measures to help address poverty, equality and decent work.
To determine
social investment
approaches and
tools for inclusive
To examine the
impact of
To bring new scientific
understanding into
company strategies
dealing with technological
content and
distribution of
Skill needs
and training
creation by
To identify
options for:
pol icy (e.g.
Fiscal policy
(e.g. robot
ta xes)
To provide ne w,
scientific insight
into technological
1.3. Purpose and structure of the guidance paper
The purpose of this guidance paper is to provide a common intellectual understanding of
BEYOND4.0. It explains what is to be analysed, why and how. As part of this task, it explains the key
developments, issues and concepts that drive the project. It should be noted that most current
policy and academic debate about the impact of digitalisation on jobs uses the term ‘future of work’.
This terminology is, we argue, simply a shorthand when jobs comprise both work and employment,
and it is this dual focus that is undertaken by BEYOND4.0 in its analysis.
After this introduction, the paper has four main sections followed by a Conclusion. Section 2 below
outlines BEYOND4.0’s focus on Industrie 4.0 and Uberisation. The empirical research design in then
explained in Section 3. Section 4 elaborates debates about technological revolutions, including the
current digital one, and the latter’s implication for work and welfare. Section 5 presents the key
concepts and issues around the four areas of empirical research enquiry for BEYOND4.0: the quality,
content, and distribution of work; the skill needs of the labour market and the skill utilisation by
organisations; the education and training to support development of these skills; and the creation
and capture (extraction) of value (and wealth) by companies. The concluding section emphasizes
how BEYOND4.0 seeks to go beyond current debates about digitalization, work and welfare, and the
need to fill the current evidence vacuum as a means to better informed policy development in the
2. Focus of BEYOND4.0 - Industrie 4.0 and Uberisation
As noted above, the main premise for BEYOND4.0 is that technology is not deterministic but socially
negotiated by key actors at various levels: firms, industry, regional, national and EU (Berting, 1993;
Bijker et al., 2012; Child, 1972; Noble, 1984). It is recognised that firms’ business models and
organisational policies and practices can differ greatly (Bloom et al., 2019; Greenan, 2003).
Company strategies are central drivers of change. They happen in the context of labour market and
welfare institutions, and public policy. Moreover, these different organisational choices shape
technological transformation as it is introduced and implemented (Kuipers et al., 2018). It is also
recognised that technological transformation is not linear but messy, occuring through economic
and political shocks that result in unpredictable socio-economic shifts (McLoughlin & Clark, 1994).
These drivers of change generally enable or disable some and not other company strategies.
Given the variety of company strategies and institutional contexts in which they are pursued, and
because some policy options are better than others (Jacobs & Mazzucato, 2016), new insights are
required about the possible range of innovative policy solutions that might be developed that
contribute to a European strategy for socially inclusive and cohesive growth and economic
BEYOND4.0 focuses on the two main technological transformations. The first is the the digitisation
of production through AI and automation/robotics, also referred to as Industrie 4.0’, with Germany
the first country to formulate policy using this terminology. The second is the digitisation of work
through the platform economy, sometimes referred to as ‘Uberisation’.
The first is based on AI combined with the emergence of big data, the internet of things and ever-
increasing computer power emabling robots to undertake both physical (manual) tasks and,
increasingly, some cognitive (mental) tasks currently performed by humans (Manyika et al., 2017;
OECD, 2018b). Although there are definitional problems, Industrie 4.0 has become emblematic of
the digitalisation of production as it emerged in Germany, applied first to manufacturing but
increasingly services (Davies, 2015; Herman et al. , 2016).
The second rests on the emergence of platform companies and the migration of work to these
platforms. Whilst different types of platforms exist, platforms are digital networks that coordinate
economic transactions usually matching the supply and demand of goods and services through
algorithms. It is perhaps the most publically visible form of digitalisation, in no small part because
of the seeming ubiquity, at least in North America and Europe, of transportation services company
Uber, to the extent that ‘Uberisation’ is suggested as the model for the future of work (see
discusison in Bernhardt, 2016).
Significantly, both of these technological transformations Industrie 4.0 and Uberisationhave the
capacity to eradicate jobs. The former by substituting jobs with technology and the later by using
technology to replace jobs with micro-tasks. In addition, both technological transformations can
also make existing skills, tax and welfare systems ineffective. Indeed, the current scientific and policy
discourse is dominated by predictions of mass unemployment, hollowed-out government and social
upheaval (see for example, Frey & Osborne, 2013; Brynjolfsson & MacAfee, 2014; Streeck, 2015).
However, the consequences of these technological transformations are still hypothetical and/or
speculative, based on econometric modelling and value prescriptions, and are as likely to be wrong
as right (Dunlop, 2016). For example, contrary to the dominant discourse about massive job losses,
many labour markets in Europe are experiencing record levels of employment. Moreover the
standard employment relationship, with full-time, permanent work for one employer, is still
dominant; though non-standard employment involving part-time, temporary or self-employment
has risen slightly (Eurofound, 2018a).
By examining the structures of power, and the conflict, compromises and acquiescences that
influence the introduction and implementation of new technologies, BEYOND4.0 seeks to generate
new scientific understanding and data on the future of jobs, business models and welfare. More
specifically, BEYOND4.0 examines the impact on social inclusion and the types of workers likely to
gain and lose in the transformation, most obviously by gender, age, education-level, industry and
location. The research also focuses on the types of jobs, and new skill, welfare and tax systems
needed to maximize the socio-economic benefits and mitigate the socio-economic costs of the new
digital technologies as well as the business models likely to deliver a fairer distribution of the gains
derived from new digital technologies. Two specific lines of inquiry are pursued. First, the
opportunities and challenges for job content, employment and productivity. Second, the
opportunities and challenges for welfare and social security systems.
Gender is an important cross-cutting issue that is embedded in analysis across the projectand for
two reasons. First, generally, women experience less employment opportunities, lower pay and
worse occupational working conditions than men (Blau & Kahn, 2017). Second, technological
transformations offer two possibilities. On the one hand, women workers may be better sheltered
from the negative impact of digitalisation because they are more likely to be employed in high-
touch, non-routinisable work that is less susceptible to automation (Muro, et al., 2019). On the
other hand, with the stress on science, technology, engineering and mathematics (STEM)
qualifications for obtaining access to and use of new technologies in companies (Lamb, 2012),
women are less represented in such educational programmes and thus at greater risk of being
excluded from the benefits of technological transformation.
3. Generating the empirical data
BEYOND4.0 uses a multidisciplinary, mixed methods research approach from which new knowledge
is generated to provide state-of-the-art understanding that that combines historical, EU-wide,
national, regional and company-level data. The project has a number of different strands of
empirical work, each with their own methods and aims. These empirical work strands often link and
are integrated into a common framework to provide intellectual and practical coherence to the
3.1. Work strands
1. Creating data for the analysis of the future of work
The aim of this strand of work is to further strengthen and integrate various relevant European
databases from which quantitative research methods are employed to generate new empirical
evidence about the socio-economic impacts of technological transformation. The activities involve
building EU cross-country and multi-level databases, covering the last decades and using existing
EU-wide harmonised surveys to allow data exploration at different levels (e.g. country, sectoral,
2. Digital transformation: regional perspectives and prospects
This strand of work identifies the current and future growth-related impacts of digital
transformation. The aim of this work strand is to identify how the EU can help support regions and
entrepreneurial ecosystems in adapting and changing course, in terms of policy pointers for an EU-
level strategy at the regional level. Both qualitative and quantitative research methods (i.e. mixed
methods) will be employed to generate new empirical evidence about the economic and social
impact in particular, the inclusive-growth related impacts- of digital transformation on six regions
and selected ecosystems within the EU. The activities will include the development of a regional
comparative analytical framework to assess the economic and social impacts of digital
transformation at the regional level, with a focus on Industrie 4.0 and digital platforms. An
ecosystems approach is used to assess the impacts along three dimensions: work and human
capital; social inclusion; and (re)distribution. A recent historical analysis of digital transformation on
inclusive growth is also conducted to assess the evolution of regional leading entrepreneurial
ecosystems in relation to digital transformation and its effects on the wider regional economy and
3. Analysing the socio-economic consequences of technological transformation
This strand of work analyses and assesses the socio-economic consequences of technological
transformation through evaluation of the existing empirical literature and datasets prepared in the
first empirical work strand outlined above. Taking into account the social embeddedness of
technology, quantitative research methods generate new empirical evidence, Europe-wide and
comparatively, on the effects of technological transformations. Using advanced statistical
techniques from econometrics and data mining, the work includes the mapping of country-level,
sector-level and company-level diversity of structural transformation, as well as forecasting the
future of work over the last two decades at the sector level and company level. Analysis also draws
upon ecosystems analysis at regional level in the second work strand above.
4. Understanding the future skills: empowering groups
This strand of work aims to provide new insights to better understand the skills needed for future
workplaces, including analysis of the skills actually used in these workplaces. Both qualitative and
quantitative research methods (i.e. mixed methods) are employed to generate new empirical
evidence and, from it, enrich policy debate on skills. Expected skills demand is identified through a
review of existing literature plus the secondary data develop in the first work strand above, the
primary data generated at the regional and company levels outlined in the second strand of work
outlined above and the sixth work stand outlined below respectively. Expected skills demand is then
compared with the supply side of vocational and education training (VET) systems and training
providers in order to clarify gaps, identify areas for possible improvements, and to provide
knowledge for better inclusion of disadvantaged employees and unemployed people.
5. Understanding technological transformations: A comparative historical perspective
This strand of work provides an historical and theoretical background to the project’s understanding
of technological revolutions and applies that understanding to policy prescription in the present.
Primarily, qualitative methods are used to generate new empirical evidence in order to understand
the effects of previous technological transformations on employment and labour markets,
particularly the cycles of unemployment and inequality, and of skilling and de-skilling.
6. Company strategies for leading economic performance and social performance
This strand of work provides state-of-the-art examples of European approaches to technical
transformation (adoption, integration, diffusion), including how stakeholders have responded to
these changes. The research also generates understanding of what elements of an inclusive
economic policy are considered important for stakeholders. Here, the focus of analysis shifts from
the regional level, as in the second work strand outlined above, to the company level. Both
qualitative and quantitative research methods (i.e. mixed methods) are employed to generate new
empirical evidence about the economic and technological effects at the company level of
technological transformations on both qualitative aspects (i.e. tasks, skills, competences,
occupational safety and health) and quantitative aspects (i.e. vacancies, job openings, types of
contracts, job mobility, polarisation, labour market target groups, gender effects) of employment.
Through a business-oriented survey and interviews an indicator is developed for this task. Results
from the analysis in the second work strand are used to identify two companies in each region to
survey both employees and HR/management, with qualitative comparative analysis (QCA) as a
research technique used to identify leading indicators and dominant business strategies. This
company level analysis also explores economic and social policies related to smart skill specialisation
and OSH issues. This information is then used to assess how the companies experience the
competitive environment both within and outside the EU. A smaller number of companies are also
selected to participate in foresight discussions and scenario building at the EU-level about the future
of work.
7. Welfare, taxes and the creation of inclusive wealth
This strand of work examines how ‘inclusive’ wealth is generated by economic processes, national
welfare systems, and taxation policies. Both qualitative and quantitative research methods (i.e.
mixed methods) are employed to generate new empirical evidence. Qualitative (desk-based)
research methods are employed to map how value created by platform economy firms can be
better captured by those who create the value (i.e. users, employers, wider public). Quantitative
research methods are used to analyse national and international (mainly OECD) tax statistics to
identify and analyse trends in income inequalities between and within countries; and between and
within sectors as well as to analyse the distribution of wealth in society. National and international
(again, mainly OECD) data is used to investigate the functional distribution of income between
capital and labour in EU Member States and the US and the impact on the concentration of wealth
in these countries, and to evaluate trends and trajectories in taxation. EUROMOD microsimulation
models are used to evaluate different policy options to extract financing for welfare state 4.0’.
Additionally, a systematic review of the advantages and disadvantages of basic income models (or
models mimicking basic income) is undertaken, including comparisons of results from experiments
in Finland, the Netherlands, Canada and some developing nations (e.g. Kenya) in order to develop
policy recommendations around the effectiveness of basic income models. A similar exercise is
undertaken on participation income models as alternatives to basic income models. Data for these
exercises is collected through interviews with managers of these experiments as well as desk-based
analysis of the welfare consequences of the models. This work strand also analyses national policies
and development trends in social security programmes to identify the possible gendering impacts
of new approaches to social policy.
3.2. Outcomes
This new empirical data generates three new outcomes for BEYOND4.0. The first is new scientific
understanding of the impact of the new digital technologies in relation to work and welfare.
Currently too much debate is framed by assumptions rather than evidence. The understanding
generated by BEYOND4.0 fills this evidence gap. The second is diagnostic and development tools to
lever technological opportunities. The third is evidence-based support for social and competitive
EU policy.
4. The Age of Revolutions?
As Figure 1 above shows, at the heart of BEYOND4.0 is technological transformation. Sometimes
this transformation constitutes a revolution, which in the case of putative ‘4th Industrial Revolution’
results in ‘digital disruption’. BEYOND4.0 recognises the potential disruptive impact of the new
digital technologies on both work and welfare. This section starts with a discussion of technological
revolutions. It then evaluates the degree to which the digital transformation can be seen as a new
technological revolution. The question is then to what degree our current policy repertoire to deal
with impacts of industrial development, is suited to deal with this revolution.
4.1. Technological revolutions
Following Schumpeter’s (1911), it has been recognised that technologies do not evolve in isolation
but are usually connected in technology systems. Every important radical innovation requires a
whole range of additional often new services, supplies and even infrastructures both up- and
down-stream. Systems are based on a combination of radical and incremental innovations, together
with organisational innovations (Freeman & Soete, 1997). Technological revolutions involve
successive technology systems. The building of a technology system sees the creation of positive
context factors or synergies, as the socio-economic context gradually adapts to facilitate the
flourishing of the new technologies. This adaptation is aided by the establishing of adequate
business arrangements and institutional context, including support facilitators such as regulations
and education (Perez, 2010).
The mass production revolution (diffusing from about 1908 to 1974) had systems around the
automobile and radio and later air travel, plastics and the so-called ‘green revolution’ (which
enormously increased agricultural productivity with successive petrochemical pesticides,
herbicides, new seeds, etc.). The current Information and Communication Technologies (ICT)
revolution (from the mid-1970s) has manifest successive systems around the microprocessor,
personal computer, internet and mobile phones and is now experiencing an evolution in artificial
intelligence, robotics and the internet of things.
Popularised by Schwab (2016) via the World Economic Forum, the notion of the ‘4th Industrial
Revolution’ is now often taken as a given and used as shorthand to describe the digitisation of the
economy and society. However, it is important to note that there are multiple ways of defining
technological revolutions. Schumpeter (1939) associated each of them with long waves in economic
growth, which would be the result of a technological revolution and the absorption of its effects.
Usefully, the neo-Schumpeterian school shifts the emphasis from ‘big bang’ innovations to a
historical sequence of successive interconnected clusters of new and dynamic inputs, processes,
products and industries, together with related organisational and institutional innovations. Such
multi-faceted analyses of technological change attempt to understand how and by what means the
new technologies diffuse and how their socio-political shaping profoundly changes our economies
and societies, not just the production processes and the jobs.
From BEYOND4.0’s perspective, a technological revolution can be defined as ‘a powerful and highly
visible cluster of new and dynamic technologies, products and industries, capable of bringing about
an upheaval in the whole fabric of the economy and of propelling a long-term upsurge of
development’ (Perez, 2002, p. 8). Moreover, a technological revolution can be characterized as ‘a
strongly interrelated constellation of technical innovations, generally including an important all-
pervasive low-cost input, often a source of energy, sometimes a crucial material, plus significant
new products and processes and a new infrastructure’, where the latter ‘usually changes the
frontier in speed and reliability of transportation and communications, while drastically reducing
their cost’ (op. cit. p.8).
This definition of a technological revolution demands a particular structure in order for a set of
complementary technologies to be seen as a ‘revolution’ (see Table 1 below). In turn, such a
structure must be assimilated by the economy and society in a complex process that requires major
policy changes, so it is a socio-institutional and indeed socio-political transformation, as well.
Table 1: The industries, infrastructures and paradigms of each technological revolution
New technologies and
new or redefined
New or redefined
Techno-economic paradigm
‘Common-sense’ innovation principles
FIRST: From 1771
The ‘Industrial
Mechanised cotton
Wrought iron
Canals and waterways
Turnpike roads
Water power (highly
improved water wheels)
Factory production; division of labour;
Productivity/ time keeping and time
Fluidity of movement (as ideal for
machines with water-power, for
transport through canals and other
waterways and for human work on
products from task to task)
Local networks
SECOND: From 1829
Age of Steam and
In Britain and
spreading to
Continent and USA
Steam engines and
machinery (made in
iron; fuelled by coal)
Iron and coal mining
(now playing a central
role in growth)*
Railway construction
Rolling stock
Steam power for many
(including textiles)
Railways (Use of steam
Universal postal service
Telegraph (mainly nationally
along railway lines)
Great ports, great depots and
worldwide sailing ships
City gas
Economies of agglomeration/ Industrial
cities/ National markets
Power centers with national networks:
decentralised centralisation
Scale as progress
Standard parts/ machine-made machines
Energy where needed (steam)
Interdependent movement (of machines
and of means of transport)
Free markets as ideal context
THIRD: From 1875
Age of Steel,
Electricity and Heavy
USA and Germany
overtaking Britain
Cheap steel (especially
Full development of
steam engine for steel
Heavy chemistry and
civil engineering
Electrical equipment
Copper and cables
Canned and bottled
Paper and packaging
Worldwide shipping in rapid
steel steamships (use of Suez
Worldwide railways (use of
cheap steel rails and bolts in
standard sizes).
Great bridges and tunnels
Worldwide Telegraph
Telephone (mainly nationally)
Electrical networks (for
illumination and industrial
Giant structures (steel)
Economies of scale of plant/ vertical
Distributed power for industry
Science as a productive force
Worldwide networks and empires
(including cartels)
Universal standardisation
Cost accounting for control and efficiency
Great scale for world market power/
‘small’ is successful, if local
FOURTH: From 1908 Mass-produced
Networks of roads, highways,
ports and airports
Mass production/mass markets
Age of Oil, the
Automobile and Mass
In USA and spreading
to Europe
Internal combustion
engine for automobiles,
transport, tractors,
airplanes, war tanks
and electricity
Home electrical
Refrigerated and frozen
Networks of oil ducts
Universal electricity (industry
and homes)
Worldwide analog
(telephone, telex and
cablegram) wire and wireless
Economies of scale (product and market
volume)/ horizontal integration
Standardisation of products
Energy intensity (oil based)
Synthetic materials
Functional specialisation/ hierarchical
Centralisation/ metropolitan centers
National powers, world agreements and
FIFTH: From 1971
Age of Information
In USA, spreading to
Europe and Asia
The information
Cheap microelectronics.
Computers, software
Control instruments
biotechnology and new
World digital
telecommunications (cable,
fibre optics, radio and
Internet/ Electronic mail and
other e-services
Multiple source, flexible use,
electricity networks
High-speed physical transport
links (by land, air and water)
Artificial intelligence,
Robotics, Internet of Things
Information-intensity (microelectronics-
based ICT)
Decentralised integration/ network
structures Platforms
Knowledge as capital / intangible value
Data as raw material
Heterogeneity, diversity, adaptability
Segmentation of markets/ proliferation
of niches
Economies of scope and specialisation
combined with scale
Globalisation/ interaction between the
global and the local
Inward and outward cooperation/
Instant contact and action / instant
global communications
* These traditional industries acquire a new role and a new dynamism when serving as the material and the fuel of
the world of railways and machinery
Source: Adapted from Perez (2002), Tables 2.2 and 2.3.
As individual technologies and systems, the diffusion of technological revolutions can be
represented as an epidemic curve, from the introduction of the first radical innovations to the
exhaustion and maturity of its potential for increasing productivity, for adding new products or
processes along the same trajectory and for expanding the market for its products and services (see
Figure 2 below).
Figure 2: The life trajectory of a technological revolution
Source: Perez, 2002
What Table 1 above also highlights is the long disagreement within academia on the periodisation
and number of technological revolutions: two, three, four, five and even six. For example,
Brynjolfsson and McAfee (2014) focus on the replacement of muscle power and brain power with
machines in The Second Machine Age, ascribing the latter to our present moment; Rifkin (2012) on
changes in energy and infrastructure in The Third Industrial Revolution and which was initially
popular with both the EU and UN. Gordon (2012) concentrates primarily on ‘inventions’ and regards
the present moment as the third ‘Industrial Revolution’.
History is an unwieldy mass of information that can be interpreted in multiple ways depending on
the analytical lens used and the notion of a technological revolution often represents a concept in
search of theoretical validation and so becomes a framework for empirical analysis. BEYOND4.0
explores these differences and their implications in detail. Nevertheless, because of its current
dominance in EU policy parlance, the starting point of BEYOND4.0 (and reflected in its project title)
is Schwab’s (2016) 4th industrial Revolution as represented below in Figure 3.
Figure 3: Schwab’s four industrial revolutions
Source: Schwab, 2016
Industrie 4.0 is grounded in the perception of there being four ‘industrial revolutions’: water and
steam power mechanisation in the late 18th century; electricity and mass production at the turn of
the 20th; microchips and computers from the late 1960s; and the emergence of ‘cyber-physical
systems’ in manufacturing today. Used foremost to analyse the consequences of technological
change on the nature of work rather than on the profound socio-economic and political dimensions
involved in such technological transformations, it is accompanied by a structuring effect of
technology as if the nature of the technologies imposed inevitable adaptation on society.
The Schwab articulation of industrial revolutions corresponds to Perez’s (2002) first three
revolutions. The second is similar to Perez’s fourth; and his third and fourth are two stages of Perez’s
fifth given that his fourth as the latest systems of the computer and automation revolution.
Significantly, for history to illuminate rather than obscure, it is important to study the full period
from the exhaustion of the previous revolution through the emergence of the new and its initial
disruptive consequences, its gradual deployment across the economy and society and finally its
exhaustion and the reiteration of the cycle. Key to this analysis is the observation of the historical
regularities in the process of social and economic assimilation of each new surge of technology:
A great surge of development is […] the process by which a technological revolution and its
paradigm propagate across the economy, leading to structural changes in production,
distribution, communication and consumption as well as to profound and qualitative
changes in society. The process evolves from small beginnings, in restricted sectors and
geographic regions, and ends up encompassing the bulk of activities in the core country or
countries and diffusing out towards further and further peripheries, depending on the
capacity of the transport and communications infrastructures. (Perez, 2002, p. 15).
This concept is different from that of long waves or Kondratiev waves (1935) as used by
Schumpeter (1939). Rather than see the process of absorption of a technological revolution as
inducing long term upswings and downswings in GDP, it focuses on the diffusion of the new
technologies and their socio-institutional absorption and shaping. This leads to different dating and
to moving away from the pure market and paying attention to the role of the socio-institutional
actors including that of government.
Each surge of development presents a regular pattern of propagation and it is these commonalities
that are the lessons that history provides. The early decades of diffusion of a revolution the
installation period are of creative destruction, led by finance. They have historically ended in a
bubble and collapse. The following period (such as the 1930s and now) reveals all the ills that
underlay the bubble prosperities, including regional and job destruction and inequality. They tend
to produce divisions in the traditional parties, between those that stick to past ideas and those that
think ahead, understanding the new circumstances. They also give rise to new movements and new
alliances, often of a populist sort, reaping the anger and resentment of the victims of the installation
period. It can be called the ‘turning point’ because it requires a change of policies and a shift of
control from finance to production, achieved by the policies of an active state. When that occurs,
the following ‘deployment’ period tends to reverse the ills and shape the technologies so there is a
positive sum game between business and society. It is in deployment that capitalism has
experienced the ‘golden ages’ such as the Victorian boom, the Belle Époque and the post-war boom.
For while the intrinsic nature of the technologies and the specific historical context are new each
time, the recurring pattern of their assimilation or lack of reveals the nature of organisations
and of behavioural habits that either resist or promote the social change processes required by each
technological revolution. The key to setting a policy course for a win-win outcome is to understand
that every previous revolution presents a wide range of potential directions. It is only by examining
the options that were available in each historical moment that the narrow view of technology
imposing an inexorable logic on society can be overcome, and policymakers and business alike
encouraged to take control of the shaping of economic growth, employment, inequality, regional
disparities, environmental problems, and all such issues that are often presented as beyond the
control of the main actors, including government.
4.2. The Digital Revolution
Digitalisation is the application or increase in the use of digital technologies by an organisation,
industry, country or region. The ‘Digital Revolution’, manifest in the 4th Industrial Revolution, is
defined as ‘a general acceleration in the pace of technical change in the economy, driven by mass
expansion of our capacity to store, process and communicate information using electronic devices’
(Eurofound, 2018b, p.1). Transforming the social organisation of economic activity, this digitisation
will impact jobs and, with it, skill, tax and welfare systems. Digitalisation is not new. Indeed, it is
based on micro-processing, the introduction and diffusion of which in the 1970s was heralded by
some at the time as a 3rd Industrial revolution (Jenkins & Sherman, 1979). What is new is its pace
and scope of technological change and its transformative potential (Meil & Kirov, 2017; Katz, et al.,
2014) – hence the claim that its drives a 4th Industrial Revolution.
The general acceleration in the pace of technological change in the economy has been driven by a
massive expansion of the capacity to store, process and communicate information using electronic
devices. New developments in robotics (cobots), Internet of Things, 3D printing, but also in the field
of big data, machine learning and artificial intelligence and the possible combinations of all, are
considered as powerful drivers for changes in living and working conditions, and mobility. The
digital transformation deeply modifies many aspects of our lives: the way we buy, sell, network,
communicate, participate, create, consume and, of course, the way we work (Meil & Kirov, 2017).
For some scholars, digitalisation has a disruptive character; for others it is a question of incremental
change or continuous evolution. For those scholars in the first camp, digitalisation opens up new
and unknown technology application potentials that will portend epochal social and economic
transformation (Avent, 2014). The counter claim from the second camp is that the new technologies
are simply like pouring ‘old’ wine into ‘new bottles’. They argue that previous debates about
automation and computerisation can inform current debates. According to them, it will be useful
to connect today’s reshaping with what has already happened and to moderate the technological
assumptions of digitalisation as a disruptive factor (Beuker et al., 2019). Despite their differences,
across both camps, there is a general consensus that digitalisation will increasingly impact work and
employment in Europe, which can have consequences for the welfare systems into which this work
and employment is embedded (Degryse, 2016; Valenduc & Vendramin, 2016; Warhurst & Hunt,
2019). However, what is also apparent is that there is an urgent need to improve the evidence base
on the actual effects of digitalisation on work and employment (Hunt et al., 2019), with a particular
weakness being the paucity of good datasets (OECD, 2018a).
4.2.1 The digitalisation of production in Industrie 4.0
The key example of the digitalisation of production is the smart industry of Industrie 4.0. Industrie
4.0 features companies using automation coupled with advanced robotics (connected to artificial
intelligence, AI) to dramatically reconfigure how goods and services are produced. AI combined with
the emergence of big data, the internet of things and ever-increasing computer power can result in
‘clever robots’ undertaking both physical (manual) tasks and, increasingly, some cognitive (mental)
tasks. Such tasks, to date, have been undertaken by humans (Manyika et al., 2017; OECD, 2018b).
These robots do not just work continuously, they are able to learn, including from machine-to-
machine information exchange, and so adapt to be more efficient at these tasks. Digitalisation thus
makes production of goods and services more efficient and more productive (World Economic
Forum, 2017).
Although there are definitional problems other labels include the ‘smart factory’ and simply
‘advanced manufacturing’ (Davies, 2015), Industrie 4.0 offers an integrated production system that,
through the new digital technologies, links not only functions within companies but also links those
companies with suppliers and customers. It offers increased production flexibility, reduced
production time, enhanced product quality and enhanced productivity. It also provides customers
opportunity to offer their own product modifications, which can then be quickly and cheaply
produced, Davies notes. It is this digital integration of conception, production and consumption
along the value chain that offers a new business model.
Whilst even in Germany the prevalence of such integrated systems is still relatively low, perhaps 20
per cent or even less of manufacturing companies (Davies, 2015), it is its potential to eradicate jobs
that has generated concern amongst policymakers in Europe. Improved computing power, AI and
robotics will replace the paid work of humans on a scale not previously seen, claims Autor (2015).
While Frey and Osborne (2013) predict that nearly half (47%) of all US jobs being at risk, estimates
vary, with some estimates lower, at around one-third (35%) of jobs (Deloitte, 2014)others are
much higher (90%) (see Lever, 2017). Underneath these headlines figures, there will be country,
regional and industry variations plus different impacts upon different types of workers (Yong Kim,
2016; Wilson, 2017; Muro et al., 2019).
Often, as Muro et al. point out, the impact upon workers is a function of regions’ industry
specialisation so, for, example, those regions with a heavy manufacturing base are more prone to
lose jobs and jobs which are typically male. The OECD (2018c) has developed a regional typology
based on employment trends and risk of jobs automation. It then identifies regions within countries
that are most at risk. The policy imperative will be to help regions at most risk by encouraging lower
risk job creation. There will also be a need to manage the risk to workers through the provision of
wrap-around welfare support by, for example, supplementing no and low incomes and ensuring the
maintenance of employability through training, education and lifelong learning to cope with more
scarce jobs.
4.2.2 The digitalisation of work in the platform economy
Part of the digitisation of work rests on the migration of work to platforms. Platforms are digital
networks that coordinate economic transactions usually matching the demand for and supply of
resources through algorithms. The use of platforms for the delivery of an increasing range of goods
and services is one of the most pervasive and visible forms of digitalisation and Uber has become
the posterchild (Walker Smith, 2016) of the putative ‘platform economy’ and ‘Uberisation’
suggested as the model for the future of work (see Bernhardt, 2016).
Such platforms can be seen to provide a more efficient or cost-effective service to citizens and
consumers (e.g. HM Government, 2015). It should be noted that different types of platform exist:
those for the exchange of goods (e.g. Ebay and Amazon Marketplace) and those for the exchange
of services (e.g. Uber and TaskRabbit). The digitisation of work centres on the latter - services.
Clients or customers purchase services, usually in the forms of a prescribed task, from a provider.
Sometimes providers bid for these tasks, sometimes they are allocated them. Services can be
routine and non-routine, local or global. Tasks can be physical (e.g. TaskRabbit), intellectual (e.g.
Upwork) and social (e.g. Bubble). Work in these platforms could be done online or mediated by the
platform, and executed offline (Meil & Kirov, 2017). Workers can register their services on the
platform and then ‘requesters’ post tasks on the platform that they want completed and an
algorithm is used to match workers to tasks based on parameters such as location, availability,
skills/features and, perhaps most importantly, user ratings. While ICT has presented this possibility
for a number of years (e.g. Amazon Mechanical Turk as a platform for micro-programming tasks
was launched in 2005) it is the integration of digitalisation that enables both a better matching of
requesters and providers the possibility of remote monitoring of work by the platforms companies
that has led to the increase in this form of working. In this sense online digital platforms represent
a new business model and also enable the delivery of services in a new way, even if some of the
services provided through the platforms are not new.
The introduction of platforms also allows the massive evolution of online or crowd employment, as
a new form of organising the outsourcing of tasks, which would normally be delegated to a single
employee, to a large pool of ‘virtual workers’ (Felstiner, 2011). In fact, crowd employment uses an
online platform to enable organisations and individuals to access an indefinite and unknown group
of other organisations and individuals to provide upon payment specific services or products
(Barnes et al., 2015). In this perspective, platforms contribute to the geographical spread of work
and employment and can create and the possibility for work to be carried out anywhere around the
world at any time in a ‘truly global, digital assembly line’ (OECD, 2017).
For workers, previously secure employment (in the case of Uber, taxi jobs) is transmuted into a
series of work tasks (with Uber, providing a client with a ride). There is an increasing consensus that
one of the most dramatic developments in European labour markets in recent years has been the
job shifts or the introduction of online platforms using ‘crowd work (Huws et al., 2016).
Digitalisation influences labour supply through the introduction of new technological intermediaries
or 'platforms' that lower barriers to labour market entry and thus include more people in the market
(European Commission, 2019a). This phenomenon alone would lead to significant disruption of
traditional industries and existing businesses, even pipeline businesses at the top of the Fortune
500 rankings (OECD, 2017). One feature of this disruption is its impact on jobs. It is platforms
companies’ potential to create new forms of work which come under considerable public scrutiny
(e.g. Taylor, 2017). However, it is not just that jobs are de-bundled into tasks, the employment
status of task providers remains unclear. Platform companies such as Uber insist that their drivers
are self-employed, refereeing to them as ‘partners’ for whom they have no responsibility to provide
health insurance, holiday pay, (paid) sickness leave or a minimum wage. Nonetheless, some
evidence suggests that the control exerted upon workers by the platform companies in terms of
how work is undertaken is reminiscent of employment (Warhurst et al., 2020f). Others argue that
neither status reflects the business model and that a new third status might be needed, such as
dependent contractor’ (Taylor, 2017) or independent worker’ (Harris & Krueger, 2015).1
Whatever its legal status, there are forecasts that growth of this type of working will continue;
Standing (2015) predicts that one third of all labour transactions will pass through online platforms
by 2025. Again, however, some argue that platforms are simply exacerbating existing non-standard
employment or atypical work that has increased since the 1970s (Cappelli & Keller, 2013). As with
Industrie 4.0, the prevalence of atypical work is still limited most workers (66%) in the EU have a
standard employment relationship (Eurofound, 2018a), characterised as full-time, permanent and
with a single employer, and engagement with platform working as a main income is low. Only
around two per cent of the adult population can be regarded as earning their main income from
platform working according to the COLLEEM survey2 of 14 EU Member States (Pesole et al., 2018).
However it remains the case that measuring the prevalence of platform work is notoriously difficult,
because it encompasses self-employment and informal work activities. A more precise view can
only be developed using a multi-pronged approach (Dhondt, 2019). Nevertheless, recognising that
the outcome for those workers experiencing atypical and gig working is precariousness, low wages,
poor prospects and social inclusion, with workers insufficiently covered by labour laws, collective
bargaining arrangements and welfare support (Kalleberg, 2009; Standing, 2011), a policy response
1 For critiques, see Eisenbrey & Mishel (2016) and Stewart & Stanford (2017).
2 The Collaborative Economy and Employment (COLLEEM) survey of 32,409 internet users undertaken in 2017 covered:
Germany, Netherlands, Spain, Finland, Slovakia, Hungary, Sweden, UK, Croatia, France, Romania, Lithuania, Italy, and
is needed, even if only to recognise the need to update social protection policies (OECD, 2018a;
EPSC, 2016).
4.2.3 Favouring inclusive and empowering technological environments
Digital technologies are not simply tools for tasks. They shape work and employment, with major
reconfigurations in the ways of doing, thinking, organising or collaborating. Thus, while digital
technologies can enhance the value of work and contribute to skills development, they can also
distort activity by interfering with operational leeway or removing from professional practices and
relationships what is meaningful to the people who do the work.
The dematerialisation of the activity can occur to the detriment of employment and work and
employee/worker wellbeing. This happens when new technologies are put in place to replace
humans or when they involve reconfigurations and requirements that destabilise work collectives,
or when goods and services resulting from them exclude some of the components of society from
the outset or do not contribute sufficiently to human development.
Thus, the development of emerging technologies must be informed by knowledge about the work
as it is really done by the workers (Kaptelinin & Nardi, 2006). Indeed, the work prescribed by an
organisation (including what is embedded in technologies), regardless of the level of formalisation
and standardisation achieved, never covers all human work. Humans keep on interpreting what is
going on in their ever-changing work environments and adjusting to it. This creation of
meaningfulness is never achieved by machines, however ‘intelligent’ they may be, since they only
retroact with their environment according to the models/expectations of the designers
‘incorporated’ into the machine. Moreover, these machines only work in addition to human
expertise that is obtained through deep knowledge and often long learning.
Taking this into account is essential in order to obtain two important outcomes from technological
development: inclusiveness and empowerment. Inclusive technological environments allow all
segments of society to become active users of the technology, in particular more vulnerable groups
as unskilled, ageing or handicapped groups. In empowering technological environments,
organisations and work situations have been designed to transform emerging technologies into
resources for work, allowing employees to flourish by performing work that is both meaningful and
efficient and thus contribute to the development of well-being and health.
Across the diversity of routes followed by companies in the digital transformation, it is important to
identify those where technologies are used as a resource complementing human work and those
where humans come as an appendix to the machines. The organisational options taken by
companies when they develop and implement technologies are crucial. They determine the
relationships between digital technologies and work content, between skill developed and skill
utilisation and hence contribute to shaping evolutions on the demand side of labour markets. New
policies are needed in order to favour the development of inclusive and empowering technological
environments in the era of digital transformation. Innovation and industrial policies would need to
be reviewed with this target in mind as well as their interactions with skills and employment policies.
4.3. The impact on welfare policy
The core of welfare policies is to provide minimum standards for the most vulnerable in society.
However, in most welfare states social policy comprises much more. In most cases social insurance
compensates the loss of income, i.e., the provisions are income-related and not targeted only to
the poor. Hence, provision can be wide: medical care, housing, pensions etc. Linked to employment,
welfare systems in each country can provide a range of support relevant to BEYOND4.0 including
education and training, unemployment assistance and income support for example. EU Member
States have their own welfare systems, which, to the point of EU enlargement, were argued to
cluster into distinct models or regimes most famously by Esping-Andersen (1990):
Liberal, with commodified, market-delivered provision;
Conservative, organised around traditional family structures and values;
Social Democratic, (high) standards based with heavily socialised provision.3
There is, nonetheless, also a European social model intended to ensure at least minimal, combining
commitments to full employment, social protection and social inclusion. Regardless of Member
State and including the European social model, Europe’s welfare systems were devised in an era
that emphasized the need for or at least desirability of, full and standard employment even in
periods of high unemployment. It was also very male-centric in focus (Wright, 2015). This approach
continues, manifest in both ‘work first’ and ‘flexicurity’ policies and practices: the first directing the
unemployed and long-term economically inactive back into work and the last is a means of ensuring
that safety nets and interventions exist to enable any necessary transitions between work. Even
whilst women (re)entered the labour market in the latter quarter of the twentieth century, often
juggling care responsibilities with employment, welfare policy was implicitly, if not explicitly, geared
to the standard employment relationship (SER).
National social protection systems developed to protect people in standard employment. This
provision is particularly the case for insurance-based schemes, i.e. those based on social
contributions from the employee and the employer. People in non-standard employment have
always been in a more insecure and precarious situation regarding access to schemes and receipt
of insurance-based benefits (ILO, 2016). Even in the most generous welfare states, those in atypical
employment are lagging behind ‘typical’ employees’ (Spasova, et al., 2017).4
In part, this positioning of welfare was governmental response to employer practice. Developed out
of the factory system, mass production involved large-scale organisations with strong internal
division of labour reliant on a consistent and dependable workforce. Attendance and performance
requirements along with the job-specific skills requirements could be achieved more effectively by
developing a stable workforce and providing permanent employment (Stanford, 2017; see also
Beynon, 1973). In this respect, as Kerr et al. (1960, pp.41-56) explained, this era of industrialism had
3 This typology has been extended and adapted, see, for example, Bonoli et al. (2000), Castles et al. (2010), and Greve (2018).
4 We acknowledge that benefits and services financed by taxes (e.g. family allowances, some forms of healthcare
and long-term care) and certain means-tested benefits (e.g. social assistance and minimum income provisions for
older people) are granted in many European countries regardless of the employment status of an individual.
a ‘logic’. It needed: a non-discriminatory, ‘open society’ in order to enable occupational and
geographic mobility within the labour market; a workforce both educated to a general level and a
specialist level related to that type of (mass production) technology and which featured both
continual training and retraining for workers; ‘structured’ workplaces and jobs ‘subject to a set of
rules’ that formalise the management and employment of workers and prevent arbitrariness; an
understanding that such rules emerge from and are administered on a ‘shared’ basis involving
government, employers and workers, what was then termed ‘tripartism’; and that, regardless, of
party-political orientation, all governments had ‘large role’ in delivering these needs.
As Stanford (2017) notes, this ‘Golden Age’ of employment was already beginning to fade in the
1970s because of macroeconomic and political changesas the rise in atypical employment from
that time identified by Cappelli and Keller (2013) illustrates. The digital revolution potentially not
just exacerbates this trend but accelerates it. If fully realised, with the eradication of human labour,
the digitalisation of production could lead to the death of many jobs, resulting in significant and
structural unemployment. Similarly, if it becomes more pervasive, the digitalisation of work will also
lead to the death of jobs, though human labour will continue but with what is now atypical work
becoming mainstreamed. Based on the SER, current welfare systems would no longer be fit for
How welfare linked to employment can be re-fitted around the new world of work is still a moot
point, though it is clear that the focus of debate is shifting. This shift is most obvious in the OECD
(2018d). In the 1990s, the OECD advocated labour market flexibility. In the 2000s it advocated full
employment. Now, worried about the existential threat to jobs, it advocates the creation of more
and better jobs and labour markets that are more inclusive and resilient. A range of suggestions for
welfare reform have been made that seek to variously remedy and prevent the death of jobs, some
of which will require ‘a fundamental rethinking’ of existing welfare provision (Annunziata &
Bourgeois, 2018, p.18). Training and education, including lifelong learning, it is suggested, needs to
be better aligned with the new skills required in the digital era, it might even be skewed more to
STEM subjects; social protection needs to be enhanced, including the provision of a universal solid
floor or safety net; workplace rights and responsibilities need to be clarified and strengthened
including legal determination of employment statuses; wages, taxes and benefits need to be
combined to protect workers rather than jobs, with benefits becoming portable, attached to
workers rather than jobs; social partnership and collective bargaining need to be made more
effective to support inclusivity and equal opportunities (e.g. Annunziata & Bourgeois, 2018; Forde
et al., 2017; OECD, 2018d; Taylor, 2017).
Kerr et al.’s (1960) logic of industrialisation was both descriptive and prescriptive. It both purported
to explain how industrialism did work and how it should work in what Schwab (2016) would describe
as the 2nd Industrial Revolution. A similar logic for the current transformation is absent but needs to
be developed. Even before the current debates about the future of work, Esping-Andersen’s (2009)
update was arguing that welfare states need to be reconfigured and made more sustainable. One
of the biggest challenges for all welfare states in the face of digitalisation is to what extent social
policy, established during the industrial era, can recognise and properly respond to new social risks,
evolving from the emerging mode of production. There is a number of alternative policy options
suggested in this context. At one end of the continuum there is universal and unconditional basic
income (UBI). At the other end, there is strongly conditional and punitive universal credit. Moreover,
there is a variety of alternative option in between these two extremes (Van Parijs & Vanderborght,
Universal basic income (UBI) is an unconditional income transfer scheme that is meant to
cover human basic needs. Sometimes these needs are defined broadly, with other income
transfer programmes becoming obsolete as result. This kind of ‘full’ basic income would
simplify the transfer system. The problem is that the full UBI would be expensive and to be
financed would require high tax rates, which might make it difficult to legitimate. An
alternative is ‘partial’ basic income. In the partial UBI the needs the benefit would cover are
more limited and the UBI would replace only the minimum level of income protection
schemes. The model would be cheaper to finance but there would be a need for other
income-related transfer schemes in addition to the partial UBI. Although the UBI is paid also
to the ‘rich’, there is a claw-back effect: the benefit is taxed away from the high-income
The idea of participatory (basic) income (PI) was advocated by the late Tony Atkinson in the
1990’s. The difference between the BI and the PI is that the PI is conditional. To receive the
basic income, people would need to be participating in society. The participation could take
different forms. It could be formal work, it could be unpaid work such as care and
volunteering in the third sector organisations, education. Thus, some kind of contribution to
society is a condition to be eligible to the PI.
It is also possible to combine the partial basic income a rather small unconditional
provision to guarantee a tight basic safety net with the PI option. Whereas the former
would be unconditional without work requirement the latter would be conditional to
different kind of participation and activities.
With negative income tax (NIT), the lowest taxable income is defined and all individuals
whose income is below that minimum level receive a ‘tax return’, i.e., they are paid negative
income tax until they will reach the accepted income level. The NIT will be gradually reduced
with increasing income form employment. All taxpayers whose income is higher than the
agreed minimum level pay income tax. NIT has some similarities with the UBI and from the
beneficiaries’ point of view the distributional consequences may be the same. However,
there are important differences. Whereas the UBI is unconditional, the NIT can either be
unconditional or conditional, and it is tested against all other income.
If the NIT would be conditional and the eligibility to benefit would be conditioned by the
claimant’s labour market behaviour or other activities, the model comes close to the
universal credit (UC). In the UC a number of basic security benefits (e.g., unemployment,
housing allowance, social assistance, etc.) are merged into a one benefit which diminishes
according to a certain percentage when income from employment increases. Those who do
not show proper activity, will lose their benefits for a period of time that can vary from a
couple of weeks to couple of years.
There are advantages and disadvantages with all the main models debated. Whereas the UBI
safeguards a solid and tight safety net, enhances freedom of choices and self-determination, there
are some problems. First, freedom of choice is a viable option only for those who have skills and
knowledge to freely choose. Second, there may be problems in combining the unconditional UBI
with activation programmes that are important for acquiring new skills needed in the digital
economy. Thus, the dilemma is to have adequate social security for all forms of employment and
simultaneously preserve incentives to work and to participate in skill-formation activities and life-
long learning. Third, it may be difficult to politically motivate the unconditional UBI. The ‘money-
for-nothing’ political frame may appear to be too persuasive. The NIT would be easier to motivate:
benefits are not paid to ‘rich’ people as in the UBI. The flip side of the NIT is to decide whether it is
an individual income or a household’s income that should be taken into consideration when defining
the amount of the NIT. Furthermore, the NIT would need an income register that is updated on
monthly basis. Conditional models demand bureaucratic screening and monitoring to establish
eligibility. The more there are conditions, the more extensive is the bureaucratic screening.
Furthermore, there is the difficult moral question whether it is right to leave those who do not
behave properly totally without social security.
In the early 2000s a new conceptthe social investment stateappeared in policy debates about
the welfare state (e.g., Esping-Andersen, et al., 2002; Morel, et al., 2012; Hemerijck, 2013; Cantillon
& Vanderbroucke, 2014). The central demand in the investment paradigm is that the emphasis in
social policy must be shifted from the ‘old’ income transfer-oriented ‘compensating welfare state’
towards a new, more proactive and preventive social investment state. In such a welfare state,
activation and capacity building are the main strategies. The European Commission (2013)5 defines
social investments as policies designed to strengthen people’s skills and capacities and support
them to participate fully in employment and social life. Thus, the social investment-based welfare
state also is an inclusive welfare state. Key policy areas include education, quality childcare,
healthcare, training, job-search assistance and rehabilitation.
An effective integration between policy areas is the crucial precondition for the social investment
welfare state. This integration ensures coherence between income transfers guaranteeing basic
safety net and income-related benefits and social, educational and employment services. The
central aim is to produce a well-educated labour force that is prepared to meet the challenges of
the emerging digital economy.
The OECD (2018d) recognises that the success of any new welfare initiatives that seek to respond
to social and economic change as an outcome of the new digital technologies will require the
building of public support. An important factor in building this support will be the generation of an
evidence base about digitalisation that goes beyond current assumptions and predictions to instead
provide data on what is happening and why and how it is happening. Only with this evidence base
can effective policy be developed. At the moment, however, that evidence base across the
digitalisation of both production and work, and at company, regional, national and EU levels is still
to be generated, and policy development hampered as a consequence.
In this respect, what Meil & Kirov (2017, p.4) state about work and welfare in the platform economy
also resonates with the coming of the robots, currently ‘there is no clear approach to the direction
that policy should be taking’ and the challenge is twofold: to identify the policy challenges and then
develop policy options. BEYOND4.0 responds to these challenges. It counters the technological
determinism that infuses much current predictions and policy about the impacts of digitalisation. It
examines not just what impacts occur to work and welfare through the putative 4th Industrial
revolution but how those changes occur and identifies the policy options that can be pursued,
particularly by the European Commission and its agencies.
5. Key concepts and issues informing the four areas of enquiry
As noted above, in examining the impact of new digital technologies, BEYOND4.0 has two areas of
enquiry for BEYOND4.0: first, the opportunities and challenges for work and employment; second,
the opportunities and challenges for welfare and social security systems. BEYOND4.0 investigates
these areas with four particular issues:
The quality, content, and distribution of work.
The skill needs of the labour market.
The education and training to support development of these skills.
The creation and capture (extraction) of value (and wealth) by companies.
5.1. This section of the Working Paper provides exposition of these issues, outlining key
concepts and issues in each case. This overview guides the further development of
the work in the different work packages of BEYOND4.0. The quality, content and
distribution of work
Creating better not just more jobs is an important goal in the EU’s Europe 2020 Strategy for
economic growth and competitiveness. Improved job quality will help address the EU’s challenges
with innovation and inclusion for example, and is positioned as a potential missing link in that
strategy (European Economic and Social Committee, 2011). This sub-section outlines job quality and
highlights some of its key features with respect to BEYOND4.0 analysis, including OSH.
5.1.1 Jobs, work and employment
Talk about the future of ‘work’ can be misleading, when the focus is jobs. Whilst ‘work’ is often used
as a catch-all term (e.g. Halford, et al., 2016), jobs comprise both work and employment. Work is
defined as any mental and/or manual activity performed by persons to produce goods and services
for own or others’ use (International Conference of Labour Statisticians, 1993 [ICSE-93]). Work can
be paid and unpaid and comprises a bundle of tasks requiring skill and knowledge. With paid work
the focus of BEYOND4.0 these tasks are put together for an employee by an employer and
applied to some form of technology. Employment is a relationship between two parties within which
the work performed by one party is paid for by the other. Employment has terms and conditions
under which that work is undertaken for the employer, and which are typically made explicit in a
contract. These contracts are governed by employment laws, regulations and guidelines, and
typically state employment status and payment. Any contract, however, is incomplete. With a right
to manage, employers have scope for different job design and employees may find themselves in
jobs that may or may not meet their needs and preferences (Knox, et al., 2015). Within this scope,
room for adjustments and trade-offs between employers and employees remain which can affect
employee engagement, performance and well-being.
The new digital technologies can impact both work and employment as well as unemployment.
Unemployment or joblessness is a situation in which a worker is actively looking for employment
but is not currently employed. In the EU some forms and level of income for some specified period
is allocated to workers in this situation through welfare and social security support.
The digital transformation may destroy businesses and jobs, displace workers and occupations,
accelerate the obsolescence of skills and hence favour unemployment. Classical schools of thought
in economics have identified several mechanisms that are likely to compensate, at least on the long
run, technologically produced unemployment. Indeed technological transformations also spur new
businesses, new occupations and skills shortages, with no certainty about that the compensating
mechanisms will offset the initial labour saving effect. Indeed, possible hindrances to compensation
mechanisms may only allow partial compensation, depending on institutional settings and on the
values of crucial parameters, such as demand elasticity, degree of competition, capital labour
substitution, demand expectations, and others (Vivarelli, 2014). Furthermore, the transformations
observed on labour markets over the last decades, challenge the definition of un-employment.
While being in employment was traditionally understood as being employed permanently on a full-
time basis, since the 1980s western economies have experienced the emergence of new forms of
employment such as temporary and often with it often short duration contracts-, part-time and
self-employment. In particular, the development of the platform economy with its digitisation of
work compounds this development, with suggestions of third status employment between
employment and self-employment.
5.1.2 How job quality is defined and measured
As yet, there is no consensus about what constitutes jobs quality or how it should be measured. A
consistent, unifying definition of job quality remains elusive (Findlay, et al., 2013). A plethora of
terms or concepts are used, for example, ‘quality of working life’, ‘meaningful work’, ‘fulfilling work’.
Each are distinct (see Warhurst et al., 2017b). For example, the ILO prefers ‘decent work’, which
tends to focus on poverty reduction in developing countries. Sometimes these terms are used
interchangeably. For example, the European Commission and its agencies use ‘decent work’ and
‘fair work’, sometimes even in the same policy document. Warhurst et al., suggest that a ‘family of
concepts’ exists that can be captured within the generic term ‘job quality’.
An addition problem is measurement, which varies by discipline, approach and form. Different
disciplines typically focus on different indicators economists favour pay for example, while
psychologists job satisfaction and sociologists tend to favour skill. The subjective and objective
traditions about measurement create further complications. The subjective approach derives
quality from the utility that a worker obtains from fulfilling his or her work. The objective approach
determines its measurement from collectively agreed amenities or dis-amenities. In terms of form,
some measures of job quality favour a single indicator, pay for example is used as a proxy in
Eurofound’s (2012) assessment of job quality trends in the EU. Currently, to measure upward
convergence of working conditions in the EU, Eurofound (2018c)6, uses seven dimensions, each
with sub-indictors: physical environment; social environment, work intensity; skills and discretion;
working time quality; prospects; and earnings .
Generating an agreed and operationalisable definition of job quality requires drawing upon and
encompassing these multi-disciplinary and multi-dimensional approaches, and which reports job
quality using an easy to understand method and which focuses solely on the work and employment
6 Having revised its conceptual framework of job quality several times over the past two decades,
that comprises jobs and omits labour market noise. Two influential attempts to do so are provided
by Muñoz de Bustillo et al (2011) and Erhel & Guergoat-Larivière (2010). The resulting six
dimensions, again with sub-indicators, developed for the QuInnE research project are: terms of
employment; pay and benefits; health and safety; the nature of work; work-life balance; and voice
and representation. These dimensions are not hugely dissimilar from those used currently by
Eurofound (2018c), in part, because they draw on the same data source, the European Working
Conditions Survey. Indeed this survey is the key data source for measuring job quality in the EU (e.g.
Green et al., 2013; Greenan et al., 2013; Greenan & Seghir, 2017; Holm & Lorenz, 2015; Holman,
2013). These QuInnE dimensions were also influential in the development of measures of ‘good
work’ in the UK following the 2017 UK Government’s Taylor Review of Modern Working Practices
(Measuring Job Quality Working Group, 2018) and underpin the new UK Working Lives Survey (CIPD,
5.1.3 The Quality of Working Life
One of the family of concepts within job quality is the Quality of Working Life (QWL). Its significance
lies in its early rejection of technological determinism and recognition that workplaces combine
human and non-human elements which best work when there is a ‘fit’ between them.
QWL emerged in the late 1940s in the context of productivity and efficiency problems in the UK coal
industry despite the introduction of new technology. The source of the problem, Trist and Bamforth
(1951) surmised from their research, was the design of work and the lack of fit between the new
technology and worker needs as humans. Their solution was to make the technical and social
systems in workplaces fit together better, with the outcome being the optimisation of
organisational efficiency.
As Warhurst, et al. (2017b) note, whilst seeking solutions in work design, QWL had an emancipatory
intent. It evolved to extend to a drive for industrial democracy and wider social change in which
good work could be the lever to a good society (Walton, 1974; Cherns, 1976). As it evolved and
became more complex it also began to suffer from definitional ambiguity (Burchell, et al., 2014). By
the mid-1970s, with macro political and economic changes, interest in it waned but did not
disappear. Interest and workplace interventions continued for example in the Netherlands and later
Flanders (Belgium) (Van Hootegem, 2016).
The Dutch variant of sociotechnical theory (modern socio-technics), stimulated by Sitter et al.
(1997), states that when organisations are confronted with an environment of increased (and
increasing) complexity, the solution is not to restore the fit with the external complexity by
increasing internal complexity. Instead the organisation should respond to this external complexity
by reducing the internal complexity. In other words, simple organisations and complex jobs. This
outcome occurs through a series of choices about structural design, job design and work processes.
In other words, whilst recognising the externalities and path dependencies, these outcomes are
largely depending on strategic choice, and managerial preferences for a top down approach versus
a bottom up approach. Another important idea in Dutch sociotechnical theory is that organisations
are not composed of two systems, technical and social systems. Organisations should not be viewed
as a balancing effort between the two systems. Instead organisations should be seen as systems
with control and operational processes. Technology and job design are means to react to
environmental demands. In this respect, digitalisation generates new requirements to which
organisations need to respond. If digital transformation is in line with market changes. In that case,
organisational design needs to follow (Kopp et al., 2019).
The original and updated socio-technical approaches offer a number of important lessons. First,
that workplaces should be regarded as sociotechnical configurations. Second, that change at work
should align with the goals of the organisation rather than be driven by technology hopes or
expectations.. Third, that innovation is an emergent practice from the interaction of the social and
the technical. Fourth that the relationship between the social and technical dimensions impact both
organisational productivity and employee wellbeing. Fifth, that the introduction and use of new
technology must be inclusive and empowering for workers. Sixth, that organisations have and make
choices about why and how new technologies are implemented, operated and extended (Badham,
2006; Kuipers et al., 2010; Oeij et al., 2017).
Interestingly, as a challenge to the current technological determinism around the digital revolution,
and the gig economy and increasing labour market precarity, there is a call for a renewed QWL
(Grote & Guest, 2017). In addition, socio-technical thinking has explicitly been applied to
digitalisation and Industrie 4.0, stressing the joint optimisation of technology, people and
organisation (Dregger et al., 2016; Hirsch-Kreinsen, 2016; Ittermann et al., 2016). Scope exists
therefore for alterative theorising about technology and organisations that is not deterministic.
5.1.4Occupational Safety and Health (OSH)
There is significant conceptual conflation and confusion between wellbeing as an outcome of job
quality and health and safety at work as a feature of job quality. BEYOND4.0 adopts this last approach
and recognises that health and safety, specially, psychosocial risks, is often overlooked within job
quality (Eurofound, 2016) and that, cast as OSH, it has an important role in delivery the successful
outcomes from digital technology.
OSH is an interdisciplinary activity concerned with the prevention of occupational risks at work.
According to the ILO, OSH encompasses the physical and psycho-social wellbeing of workers. OSH
research needs to address a wide range of potential problems (Bundesanstalt für Arbeitsschutz und
Arbeitsmedizin, 2010; EC, 2009) from mechanical, physical, thermal and biological hazards through
to fire and explosion risks to physical working environment factors to physical strain (e.g. workload,
lifting, awkward postures etc.) through to psychosocial hazards (e.g. work design and work
organisation, and social conditions).7
A distinction can be made between work-related disease and occupational health. The concept of
a work-related disease includes diseases in which work plays a role. The concept of a work-
aggravated disease is one which is made worse by work, regardless of the original cause (Eurostat,
2004). The EU LFS ad hoc module of 2007 lists respondents’ most serious work-related health
problems (Eurostat, 2010): musculoskeletal disorders; stress, depression or anxiety; breathing or
7 ILO Occupational Safety and Health Convention (C155), 1981; Resolution concerning statistics of occupational injuries (resulting
from occupational accidents) adopted by the 16th International Conference of Labour Statisticians, 1998; European Commission,
European Statistics on Accidents at Work (ESAW) Methodology, 2001; European Council, ‘Council Directive 89/391/EEC of 12
June 1989 on the introduction of measures to encourage improvements in the safety and health of workers at work (Framework
Directive)’, Official Journal L 183, 29/06/1989, 1989, pp. 0001-0008.
lung problems; heart disease or attack, or other problems in the circulatory system; headache
and/or eyestrain; infectious disease; hearing and skin problems. Occupational health principally
means the absence of occupational diseases (Eurostat, 2010). However the WHO adopts a wider
perspective, defining health as ‘a state of complete physical, mental, and social well-being and not
merely the absence of disease or infirmity’.8 This approach aligns with the joint ILO/WHO
Committee’s definition (Pintelon & Muchiri, 2009, p.613):
Occupational health should aim at: the promotion and maintenance of the highest degree
of physical, mental and social well-being of workers in all occupations; the prevention
amongst workers of departures from health caused by their working conditions; the
protection of workers in their employment from risks resulting from factors adverse to
health; the placing and maintenance of the worker in an occupational environment adapted
to his physiological and psychological capabilities; and, to summarise: the adaptation of work
to man and of each man to his job.
While it is difficult to define, safety means free from harm or risk. In practice this state is impossible
to achieve. As a consequence, safety must be seen as a value judgment regarding the level of risk
of being injured which is considered to be acceptable (Harms-Ringdahl, 2001). In this respect, safety
is traditionally seen as accident prevention but can also be seen as a basic value in the workplace.
From an employer perspective, poor OSH can be detrimental to organisational performance.
Evidence shows that stress at work leads to an increase in accidents (Clarke, 2010), longer periods
of sickness absence and greater staff turnover (Coomber & Barriball, 2007). The combination of
high job demands and low control has been shown to lead to mental health problems (e.g. Madsen
et al., 2017). Musculoskeletal problems can lead to workplace absence (Duijts et al., 2007; Steensma
et al., 2005) and even labour market withdrawal into disability benefits (e.g. Canivet et al., 2013) or
to coronary heart and vascular disease which may result in hospital admission or mortality (e.g.
Kivimaki et al., 2012) all of which, of course, are also detrimental to workers and also creates
health and welfare costs for EU Member States Alli’s (2008) point about the need to adapt work to
workers and adapt workers to his or her job in the context of digitalisation is therefore crucial.
Indeed, discussion of OSH has changed because of the technological transformations. High-tech
systems hold the lure of preventing all accidents and creating inherent safe work situations
(Parmiggiani et al., 2014). It seems that the goal of Zero Accidents is finally achievable, provided by
the robots and cyber-physical systems (Reardon et al., 2015; Zwetsloot et al., 2017). Experiments in
OSH in the field of technology are driven by the idea that inherently safe robot systems are possible,
with technology solutions that prevent harmful contact scenarios occurring with operators
(Guiochet, 2016) and OSH regulations and the EU Machinery directive seeking to ensure ‘risk-free’
design. However, in practice design choices are technology driven and do not include the OSH
analysis from the operator perspective. BEYOND4.0 identifies the health and safety approaches used
in Industrie 4.0 and Uberised work settings. Resonating with the socio-technical systems approach,
BEYOND4.0 seeks to develop a new, more operator-centred OSH approach to cyber-physical
8 WHO (1946) WHO definition of Health, Preamble to the Constitution of the World Health Organization as adopted by the
International Health Conference, New York, by the representatives of 61 States (Official Records of the World Health
Organization, no. 2, p. 100) and entered into force in 1948.
systems based on new approaches such as systems thinking (Leveson, 2011) and the Zero Accident
Vision developed by TNO (Zwetsloot et al., 2017).
5.1.5The distribution of work
The point about the importance of the inclusiveness of technology resonates with EU policy
concerns. The technical division of labour relates to the fragmentation of the production process
into specialised or at least discrete tasks undertaken by different workers. Relatedly, the division of
work is sometimes used to refer to the division of a large task, contract or project into smaller tasks.
It is increasingly recognised that it is these tasks that can colonised by the robots (Industrie 4.0) or
allocated as gig work (Uberisation) and so affect the distribution of future work. Whilst worker
allocation to these tasks is, in theory, predicated on specialized skills and therefore assumption
about their productive capacity, in reality, sorting by type of workers is influenced by other factors
such as sex, race and skill possession.
In microeconomic terms, the distribution of work results from the production process and more
precisely, from how the inputs relate with one another in the production function. Two concepts
are important in the economic analysis of how technological change interferes with productive
interdependencies: capital-skill complementarity and technological bias. Two factors are
complementary: when the quantity used by one increases then the price of the other decreases.
The partial substitution elasticity is, in this case, negative. Griliches (1969) has analysed capital-skill
complementarity in order to explain the continued growth in the relative wages of skilled workers
in the US despite the growth in the supply of this category of labour. He carried out an empirical
test on a cross-section of sectors for 1954 that confirmed the hypothesis of complementarity
between capital and skilled labour. Subsequently, a large number of studies have confirmed the
Griliches result. According to Hamermesh's synthesis (1993), the partial elasticity of substitution
between skilled labour and capital is generally negative; the partial elasticity of substitution
between capital and unskilled labour is always positive and there is no consensus on the partial
elasticity of substitution between qualification categories. A technological bias is a distortion in
factor shares in the presence of technological change. When a capital-skill complementarity is a
property of a given production function, a technological bias relates to shift in the production
function due to an external shock.
Using French company-level data for 1986-1991, Duguet and Greenan (1997) find an overall
innovation bias favouring skilled labour. A second effect reinforces the former: unskilled labour is a
stronger substitute to capital than skilled labour. However, they also show that not all innovation
types reduce the unskilled labour share. Incremental product and process innovations tend to
increase it. The skill bias results from the aggregation of several impacts depending on innovation
types. In total, the distortion of factor shares in the presence of technological progress depends on
productive interdependencies that are very seldom looked at precisely in empirical studies. This is
because not many datasets provide direct measures of technological progress as well as measures
of factor shares and costs but which would be needed to identify what is happening in the
production function. Second, the way firms innovate, and in particular whether this innovation is
incremental or radical and whether it encompasses organisational changes, probably generates
different regimes of productive interdependencies.
The concept of technological complementarities exposed in Brynjolfsson and Milgrom (2013)
identifies another type of productive interdependencies. Technological complementarities refer to
a discrete choice organisational design function. When there is a complementarity between several
pairs or organisational design decisions, then these decisions should cluster. This occurs is because
when a set of organisational features or practices are complements, implementing then together
yields higher productivity gains than implementing each of them separately. The existence of
productive complementarities therefore encourages radical innovation since organisations will
benefit more from a joint change in multiple dimensions than from a partial implementation. At the
beginning of the millennium, the management science and economic literature has highlighted the
complementarity between the adoption of ICT and the implementation of changes in work
organisation (Ennen & Richter, 2010; Bresnahan et al., 2002). The assessment of productivity
impacts of digital transformations should rely on a production function augmented by an
organisational design function. Such a frame is likely to explain the great heterogeneity in the
productivity impacts of technological transformations.
A recent empirical study by Hunt et al. (2019) sheds light on the effects of AI with automation in UK
organisations. It shows that the occupations most impacted are professional and higher technical
staff, and managers, administrators and intermediate managerial staff. Whilst job elimination, job
creation and job change occurred across the organisation, the jobs eliminated by AI substituting for
physical tasks tended to be at the low-skill level, with jobs created at a range of skill levels. By
contrast, AI substituting for cognitive tasks tended to eliminate jobs at a range of levels and create
jobs at the high-skill level. Overall a general upskilling of tasks was therefore apparent, with
employees needing more skills and knowledge in three-fifths (60%) of organisations introducing AI.
If this trend holds for other countries, it would mean that employers in the future are likely to
require more higher-skilled workers. This research also showed that the introduction of AI for
physical tasks was more common in organisations where the workforce was mainly low skilled. The
introduction of AI for cognitive tasks was more common in organisations with a predominantly
younger or high-skilled workforce. Adoption of both of these forms of new technology was lowest
among organisations that had a mostly female or a mostly older workforce. A deeper understanding
of productive complementarities associated with digital transformation would give stronger
foundations to the analysis of structural changes in workforce composition.
Similarly, US industry level rather than organisational level analysis reveals differential impacts by
types of worker. Male workers appear more vulnerable to technological unemployment because
they work more in manufacturing and transport, whilst women are safer because of their clustering
in the health and education sectors. Hispanic and black workers in the USA are more vulnerable
than white or Asian workers due to Hispanic workers’ over-representation in the construction and
agricultural sectors and black workers in transport (Muro et al., 2019).
Other EU-wide research (Hunt et al., 2018) has attempted to examine employment participation
rate outcomes for the type of workers identified as vulnerable in the labour market female, older,
younger, migrant and low-skilled workers (EC, 2010) in EU countries classified as high on
technological innovation. The findings show that reduced inequality, as measured by higher
employment participation, is not comprehensive for vulnerable workers within a high innovation
regime. Indeed, there is no clear evidence that high innovation can be expected to inevitably reduce
inequality for these workers. However, there are again variations. For example, compared to non-
vulnerable workers that is, male, prime age, middle and high skilled and native born employment
participation for low-skilled workers is worse but better for female, older and migrant workers. It is
vital therefore that any analysis of the impact of digitalisation is sensitive to the uneven distribution
of work that might follow and whether it compounds, improves or simply continues existing labour
market inequalities and exclusions.
5.2. The skill needs of the labour market
Any understanding of the future of work needs to examine the impact of the new digital technology
on skills, and how the supply and demand for skills, particularly digital and specialist IT skills, are
changing. Indeed, the European Commission believes that ‘the future of work is all about skills’
(EPSC, 2016, p.7). This sub-section defines skills, indicates what skills might be needed and the
important issue of ensuring that supply meets demand in the labour market.
5.2.1 Defining skill
A task is a unit of work activity that produces output (Autor, 2013). All paid (and unpaid) work
comprises a bundle of tasks: physical, intellectual and social (Fernandez-Macias et al., 2016). The
balance of these tasks can vary by job but each of these tasks is underpinned by skills and
knowledge. Whilst skill and knowledge can be conceptually separated, in practice they can be hard
to disentangle, particularly if the exercise of skill requires knowledgeable practice (Thompson,
Skill level is assessed through the complexity and range of tasks associated with a job and skill
specialisation associated with the field of knowledge required, the tools and machinery being used,
the material being worked on or with and the kinds of goods and services being produced (Elias and
Day, 2017). With respect to specialisation, skills are often designated as domain-general and
domain-specific. The former skills are transversal across occupations, the latter confined to
particular occupations. For example, the former can include punctuality, the latter technical
capabilities such as being able to write a computer programme. At this point, however, any potential
scientific or policy consensus on what constitutes a ‘skill’ quickly evaporates. Punctuality, to return
to that example, could be regarded as a skill or a personality trait (Grugulis et al., 2004) and, if
regraded as a skill, might only be so for political reasons in the power play between employers and
unions over pay (Warhurst et al., 2017a).
To compound the definitional problem, understanding of skill is dynamic, changing over time
(Grugulis et al., 2004; Warhurst et al., 2017a). To return again to the above example of punctuality,
what might once have been regarded as personality traits can now be cast as skills. It can also be
spatially specific. In some countries, ‘skill’ still refers to having and being able to apply accredited
vocational knowledge acquired through a mixture of formal and on-the-job learning. In other
countries it now means whatever employers want it to mean (Lafer, 2004; Warhurst et al., 2017a)
However, if tasks are underpinned by skills then it is possible to suggest that there are technical,
behavioural/social, cognitive and basic skills (Mournier, 2001; Green, 2011), with the last of these
skills consisting of reading, writing and computer literacy skills (EC, 2019b).
To complicate matters further, the concept of ‘T-shaped’ skills is a metaphor used in job recruitment
to describe the abilities of workers. The vertical bar on the T represents the depth of related skills
and expertise in a single domain, whereas the horizontal bar is the ability to collaborate across
domains with experts in other areas and to apply knowledge in areas of expertise other than that
of the particular workers. Other shapes have also been proposed, for example ‘X-shaped’
(leadership), ‘I-shaped’ (individual skill depth without communication skills) and ‘Tree-shaped’. The
tree-shaped worker has more rhizomic skills with depth in many areas, not just one, and being able
to reach many heights of accomplishments in many different fields or many different branches of a
domain. Finally, Gamma- (Γ) and Mu- (Μ) shaped individuals have been described by Fiore-Gartland
and Tanweer (2018) based on ethnographic research of data science research communities and
which characterises workers with supporting strengths in computationally- and software-intensive
Being broadly and variously defined, skills consequently lack common measurement internationally
(Cedefop, 2017). In the absence of definitional consensus, what gets counted as a skill is that which
can be measured skills that are credentialised with qualifications (Felstead et al., 2017) and
become so-called ‘hard’ skills as opposed to soft skills, which, in the melee of types of skill have
come to constitute everything else. One consequence is that qualifications and skills tend to be
conflated, even treated as synonymous.
Whist accepting the problems with definitions and measurement, there are two key reasons for
continuing to attempt to focus on skill. First, skills (in the form of qualifications), along with the
training needed to acquire those skills, characterise occupations and, indeed, are used, to
hierarchise occupations in occupational classifications systems e.g. ISCO. If new jobs with new tasks
emerge, workers will need the skills to perform these tasks. Understanding these skills is important
so that, at the very least, appropriate education and training provision can be designed and
delivered. Second, whilst debate about the future of work is currently dominated by the death of
jobs, past evidence (e.g. Levy and Murnane, 2004) suggests that it is equally likely that the tasks and
so underpinning skills of jobs will change as much as jobs will disappear and emerge. How the
balance of skills within those residual jobs changes will thus equally need to be understood. Indeed,
there is a discernible turn in debate about the future of work to acknowledge that, along with job
destruction and creation, tasks will change within existing jobs (Eurofound, 2016).
5.2.2 Skills for the future
In terms of what skills will be needed in the future, comprehensive empirical data is still lacking on
the impact of digitalisation on EU jobs. This problem is compounded by skill being poorly analysed
and conceived in current debates about the digital economy. The problem starts with Frey and
Osborne’s (2013) influential calculation of US future skills needs based on occupational level
analysis. They expect many occupations to disappear. However, occupation is not the right
observation level for changes in the labour market. Atkinson and Wu (2017) point out that
‘occupational churn’ that is, the constant creation and destruction of jobs is at an historic low in
the US.
Moreover, Frey and Osborne (2013) treat jobs as non-replenishable goods. However, jobs change
constantly, and that it is important to create environments that help job occupants shape their jobs
and have jobs that improve their skill sets. This is called the ‘AMO’ approach; workers having the
‘ability’, ‘motivation’ and ‘opportunity’ to expend the discretionary effort that supports company
innovation and competitiveness (Appelbaum et al., 2000).
Another weakness is that Frey and Osborne’s analysis centres on a crude distinction between
routine and non-routine tasks within occupations (Pfeiffer, 2016). Another approach recognises
that the experience of employees plays a dynamic role in shaping tasks within jobs. The operator of
the future needs to have knowledge of many topics and knowledge areas to make appropriate
interventions in high-tech environments. Both criticisms fundamentally shift the focus of analysis
onto task- rather than occupation-based approaches for estimating the impact of digitation on skills.
Pfeiffer proposes an index based on labouring capacity to describe automation-resistant
components of human work action as a multi-dimensional interplay of complex challenges in
specific work situations and the action dimensions necessary for adequately responding to these
challenges. However, progress has stalled because a thorough analysis of the connection between
both is missing.
In the absence of good data, debate exists about how new technologies in general shape the stock
of skill (see Hunt, at al. 2018). On the one hand, the theory of Routine Biased Technological Change
argues that technological change in production processes9 leads instead to job polarisation, with
decreased demand for workers with intermediate skills and relative increases in low and high-skilled
occupations. With digitalisation, routinisable, intermediate technical (Industrie 4.0) and managerial
(Uberisation) skills can be replaced by robots and algorithms. These intermediate jobs then
disappear, substituted by technology. The stock of remaining or new jobs then polarises, with the
expansion at the top and bottom of the occupational hierarchy of high-touch service jobs that are
non-substitutable by technology (Appelbaum, 2012) as well as, at the top of the hierarchy, jobs
growth in higher-skilled jobs that require complex skills (see also Goos et al., 2009, 2014; Acemoglu
and Autor, 2010). On the other hand, the theory of Skill Biased Technological Change argues that
technological change tends to increase demand for skilled workers and decrease demand for low
and unskilled workers (Levy and Murnane, 1992, 2013; Violante, 2008). Digital technology will thus
replace lower-skilled jobs on the one hand, while creating high-skilled occupations on the other.
Within the New Skills Agenda for Europe stress is placed on developing ‘digital skills’, though the
nomenclature changes depending upon the discourse: ‘21st century skills’ and ‘T-shaped skills’
being the most obvious. The main argument in the New Skills Agenda for Europe is that digitisation
requires a higher level of skills than ten years ago. What these skills are in practice is difficult to
discern and there is need for European Frameworks (e.g. EQF, ECVET, EuroPass, EQUARF) to be able
to make these skills transparent and comparable to enable appropriate education and training. EU-
wide recognised skills and occupational profiles need to be refined. Doing so reduces existing skill
mismatches between labour demand and supply. This is a particularly important task given that
mismatches have higher incidence in polarised labour markets in Europe (Sarkar, 2017).
Notwithstanding this need, the anticipated result is an upskilling of the workforce rather than
polarisation. In this respect, there is a strong policy emphasis currently on STEM. The US
Department of Labour, for example, has identified 14 industries with projected jobs growth or which
will or will affect the growth of other industries or are being transformed by technology).
Notwithstanding the lack of consensus about what constitutes a STEM occupation (Elias and Day,
2017), it is assumed that STEM skills will be crucial for what are called the Key Enabling Technologies.
9 Supplemented by expanded opportunity to offshore labour also enabled by technology.
In the US, the Committee on STEM Education of the National Science and Technology Council has a
five-year strategic plan to raise investment in STEM-related technical education programmes.10
Similarly, in the EU, the High-Level Expert Group on the Impact of the Digital Transformation of EU
Labour Markets (EC, 2019a) warned against creating polarised labour markets and urged the
European Commission to ensure that it encouraged inclusive labour markets by the promotion of
digital skills. A dual approach results: the promotion of basic digital literacy skills to address current
deficits in these skills amongst some segments of the workforce, and promotion of advanced digital
skills to plug potential skills shortages in industries with high performance computing needs (EC,
Part of any modernisation of VET systems within the EU requires recognition that, whilst important,
workers having digital skills alone will be insufficient. There are digital skill deficits amongst some
workers. However, there are other skill needs too. The different tasks identified by Fernández-
Macías, et al. (2016) map onto different skill sets, all of which are relevant to the digital economy.
As a consequence, whilst working with the New Skills Agenda for Europe, BEYOND4.0 seeks to refine
it, widening the perspective towards the range of skills requirements that foster innovation and
adaptability (Barnes et al., 2016; Howaldt & Hochgerner, 2018). These skills might be job-specific
and/or generic and acquired through formal and/or informal learning processes (James et al., 2013).
A re-assessment of the basic skills-measurement approaches is needed. BEYOND4.0 builds on the
suggestions of Pfeiffer (2016) and others to offer a new assessment of future skills needed to help
workers gain, maintain and progress within employment. Importantly, this assessment will explicitly
make distinctions between the different levels of skills needed from the unemployed through to
management levels. To do so effectively again requires that the occupation perspective be
connected to the task and the organisational level. Companies need to know how to help individuals
expand their skillset during their working career. This skill acquisition requires companies not
reducing work to specialized small tasks but creating real T-shaped organisations. Team
environments are needed that integrate individuals with overlapping high-tech skill profiles (Dhondt
& Van Hootegem, 2015).
5.2.3 Skills mismatch
The point about wanting to avoid skill shortages in the labour market as the future of work unfolds
looms large in EU policy thinking. These shortages are one form of skills mismatch or imbalance
(Gambin et al., 2016; McGuinness et al., 2018). The OECD (2015) defines a skill mismatch as sub-
optimal allocation of workers to jobs resulting in over- or under-qualification. This definition uses
qualifications as the proxy of skills and measures skill mismatch at the level of the employee level.
When measured at the employee level, skill mismatch refers to the extent to which the skill level
possessed by a worker does not match the skill required to do the job. It is also referred to as vertical
skill mismatch. Horizontal mismatch refers to a mismatch between the employee’s field of study
and field of study required in the occupation. When the skill level possessed by a worker
corresponds to skill level required by job, the worker is well-matched. When skill level possessed by
worker is higher than that required by job, the worker is over-skilled and a skill surplus exists. Finally,
when the skill level possessed by a worker is weaker than that required by job, the worker is under-
skilled and a skill deficit exists, see Figure 4 below.
Figure 4: The impact of skills imbalances at an employee and employer level
Source: McGuiness et al. (2018)
As Figure 4 also shows, mismatches however can occur for organisations in relation to both internal
and external labour markets. The first is called a skills gap and refers to a situation in which an
employer believes that existing employees do not possess the skills to successfully perform their
tasks. The second is called a skills shortage and refers to aggregate supply in the labour not meeting
demand, and is manifest for employers in recruitment problems due to a lack of suitably qualified
candidates (Gambin et al., 2016; McGuiness et al., 2018).
The concept of skill mismatch thus requires measures of both the skills possessed by workers and
skills required in work. Caution however is needed at this juncture for two reasons. First, because
employers can, and often do, complain of skill shortages whilst scaling back their own workforce
training provision and can have recruitment difficulties not because too few suitably skilled workers
exist but because the reward package offered is unattractive to suitable skilled workers. Hard-to-fill
vacancies then exist rather than pure skills shortages (Gambin et al., 2016). Second, because
employer skill demand has two facets: the skills workers are required to possess in order to get a
job and skills to do a job (Warhurst and Findlay 2012). As Warhurst and Luchinskaya (2018) reveal,
most analyses are driven by data availability and most skill surveys focus on the first type of demand
but the two are not always the same. Investment in human capital, in the form of higher education,
has expanded in the developed economies in anticipation initially of emergent knowledge
economies and now the emergent digital economy. Faced with a supply of better-qualified
applicants, employers simply hire workers with better qualifications, viewing the possession of
qualifications as a signal of capability (Rothschild & Stiglitz, 1976). However, this employer
behaviour leads to ‘qualification inflation’ with the skilled (proxied by qualifications) need to get a
job exceeding the skills need to do the job. The problem, as James et al. (2013) note, is that, for a
number of reasons, employers might not recognise or deploy these higher-level skills resulting in
workers having a skill surplus and being under-utilised in work. In this respect, skill surveys are both
Skills imbalances
Employee level
Vertical mismatch
Skills mismatch (over-
Horizontal mismatch
Employer level
Skills gaps
Skills shortages
valuable and limiting and mixed methods are useful in generating a more rounded understanding
of employer skill demands a point manifest in BEYOND4.0’s research design.
This point is important. With a rapid expansion of higher education provision in the advanced
economies, there are concerns that a structural over-supply of skilled workers now exists and that
too many workers’ skills are under-used in their work. This skill under-utilisation matters for both
forms and their employees for four inter-related reasons. First, it means that firms are not
maximising the human resources available (OECD, 2011). In this respect, Livingston (2017) identifies
skill under-utilisation as a cause of under-performance for firms. Second, it creates employee
disaffection, with consequent employee recruitment and retention problems (Skills Australia, 2012;
Okay-Somerville & Scholarios, 2013). Third, for employees, if firms are under-performing as a result
of skill under-use, it reduces the possibility of wage mobility. Firms with higher productivity still
typically pay higher wages (Bosworth et al., 2019). Fourth, and also related to the sub-section below,
it can disincentivize workers form investing in skill development when they are increasingly being
asked to take on the financial risk of investing in that development (Gambin & Hogarth, 2017)
whilst at the same time digital transformation will require new or at least reconfigures skills amongst
the workforce and potential skill shortages are signalled. If the digital transformation is to be
inclusive then ensuring effective skill use needs to be a priority for Europe.
5.3. Education and training to support skill development
This sub-section examines vocational education and training (VET) to support the development of
skills deemed necessary in the digital economy and to sustain employability and enhance social
mobility amongst EU workers. It outlines the importance of VET and the skills systems that deliver
VET. It also highlights the important role of lifelong learning in the development of the skills that are
needed before raising a number of general and digital-economy specific challenges to
(re)configuring VET in the EU.
5.3.1 The importance of vocational education and training
VET systems need to be able to provide a skilled labour force prepared for the digital
transformation. The New Skills Agenda for Europe already includes measures to support the
modernisation of VET. Whilst responsibility for skill development lies with Member States, the
European Commission has identified the need for change in the way skill systems adapt to the digital
revolution (EC, 2015a & b). Demand is growing for digitally-skilled workers in Europe and, as noted
above, there is concern about future skills shortages unless remedial action is taken. The European
Commission has called for digital skill levels to be raised among employees in all industries and
amongst job seekers in order to improve their employability (EC, 2015a & b). However these digital
skills will need to be complemented by both a higher level and a broader range of skills to ensure
that employability is maximised. While empirical evidence is currently lacking, there is a concern
that digitalisation will widen existing inequalities, creating or at least compounding job polarisation
in the labour market (Autor et al., 2015). Skills development must therefore address both
employability and social mobility if it is to ensure that workers can gain, sustain and progress in
employment, and in addressing progression opportunities, avoid any potential ‘bad jobs trap’ that
might create political disaffection within the EU and towards the European project (Eurofound,
As Behle (2016) notes, the concept of employability has evolved. At the start of the twentieth
century, it centred on a duality of being employed and being unemployed. By mid-century it had
shifted to focus on supply-side factors, principally the skills required to be employed, Behle states.
However what is regarded as skills for employability can vary by country and their transferability
across industries can be over-stated (Wheelahan, 2017). Now, it is recognised that employability
involves analysis of demand not just supply-side factors and, so, most obviously employer skill
requirements and the state of the labour market generally (Nickson et al., 2012). Similarly, there is
debate about what constitutes social mobility and how it should be measured. Economists tend to
use income (e.g. Blanden et al., 2005) while sociologists tend to focus on class (e.g. Goldthorpe &
Mills, 2008). Typically in the first approach, movement in an individual’s wage levels constitutes
mobility; in the second approach, intra- or inter-generational occupational movement denotes
mobility. In each case, however, skills are important in enabling access to higher level jobs with
better pay. The education and training that delivers these skills consequently has demonstrable
effects on labour market inclusion and social mobility (as well as productivity and innovation, see
5.3.2 Skill systems
Skill development or ‘formation’ systems involve a number of institutions; the state, the market,
firms, business associations and, outwith these formal institutions, a number of informal ones such
as communities (of practice), clans, clubs and networks (Crouch, 2006) to which should be added
families and friends (James et al., 2013). The state has two interests in skill development: first, a
general responsibility for ensuring for basic and advanced levels of education for their citizens;
second, as a means of underpinning, if not driving, economic development in which case the
emphasis is on vocational skills. Ultimately however it is individuals who need to acquire and possess
these skills and employers who require them, either ate the point of hire of the point of use..
If once company product market and national business systems were regarded as drivers of skill
development, the importance of negotiated or at least mediated choices by key actors is
increasingly being recognised as a factor (Ashton et al., 2017; Warhurst and Luchinskaya, 2017). In
this respect, skill development systems are not determined by technology rather they consist of
configurations of institutions representing ‘social settlements’ between civil society (employers,
labour and occupational groups), the state and VET providers (Domanski & Kaletka, 2018). Each
national system is therefore the outcome of consensus, compromise and conflict, and within each
national settlement the balance of power in the relationships between civil society, state and
providers varies (Bosch, 2017; Wheelahan, 2017).
Skills development can also vary: skills can be acquired in different ways and acquired at different
life stages, work tasks can be vertically or horizontally distributed and organisations can modify their
skill needs through work design, including technology use. Actions related to each of these four
strategies come with costs and benefits, and strategic choices are made (Keep, 2017).
Moreover, although much of the skills system analysis focuses on formalised VET provided through
qualification awarding bodies, informal workplace leaning also matters because it links with
workplace innovation (Lundvall, et al., 2008). Workplace innovation regards employee engagement
as a condition for successful technological innovation that simultaneously supports business
performance and good quality of work; in other words it seeks the joint optimisation of business
and employees. For Oeij and Dhondt (2017, p. 66), workplace innovation is defined as ‘an integral
set of participative mechanisms for interventions relating structural (e.g. organisational design) and
cultural aspects (e.g. leadership, coordination and organisational behaviour) of the organisation and
its people with the objective to simultaneously improve the conditions for the performance (i.e.,
productivity, innovation, product quality) and the quality of working life (broadly defined as
wellbeing at work.
What calls for workplace innovation also underline is that much current thinking about skills
acquisition is premised on a ‘unitary’ assumption that there exists opportunity to deploy these skills
in work an assumption that is problematical empirically (Wheelahan, 2017, p.644). Workplace
innovation signals that the benefit of skill development can only be realised in work environments
that offer that opportunity and also motivation for their exercise the Ability-Motivation-
Opportunity (AMO) framework signalled as creating more productive workplaces by Appelbaum et
al. (2000; and of which more below). Preventing a biased attention for HR-related interventions,
the basis of skill development is, according to workplace innovation, rooted in the structural design
of organisations and jobs. Structural design is a root cause for good quality of work, whereas HR-
related interventions merely mitigate negative effects of such design (Oeij & Dhondt, 2017).
Encouraging organisations to adopt these organisation and management approaches requires
locating the skill system within a wider ‘skill ecosystem’ consisting of a range of interdependent
political and economic institutions and which has adaptive capacity to changing conditions and can
be directed to delivering more progressive and shared skill outcomes. Importantly, there is an
emphasis on system-wide capacity-building to plan and manage VET, with stakeholders recognising
and committed to a broad agenda of individual, business and local economic interests (Anderson &
Warhurst, 2012).
The skill ecosystem approach is more complex and dynamic than the ‘one-stop, quick-fix’ supply
side focused thinking of much current policy (Anderson & Warhurst 2012, p.119). It requires a more
and closer interaction between the different institutions and actors in the relevant sector or
regional ecosystem (Warhurst, 2017). It takes understanding of VET needs for the future of work
beyond the usual market or state dichotomy and in the direction of a coordinated approach that
might result in a new skill settlement.
5.3.3 Lifelong learning
The call for higher level skills to overcome the future employment disruption caused by digital
transformation also connects to renewed discussion about the role of life-long learning to help
foster workers’ adaptability to changing labour markets over their working life (Barnes et al., 2016).
The European Commission (EPSC, 2016) believes that continuous skill development best guarantees
life-long resilience in the uncertainties surrounding the future of work. Key to this continuous skill
development is lifelong learning and adult education. The idea that education should be extended
throughout life was inspired in the 1970s by UNESCO under the slogan of lifelong education. In the
mid-1990s the concept of lifelong education gave way to the concept of lifelong learning. While the
paradigm of lifelong education assigns a central role to the state in ensuring wider educational
opportunities for all citizens, the lifelong learning paradigm shifts this duty mainly to the individual.
It is a multidimensional concept which refers to different kinds of knowledge and skills (formal and
informal, planned and ad hoc, purposeful and unintended) within different perspectives (personal
and social; employment and citizenship; leisure and work). Adult education is the most important
form of lifelong learning.
Lifelong learning refers to a radical and all-encompassing change in education and learning, which
implies a quantitative growth of education and training opportunities, a constant rethinking of the
contents of education and training activities, the unfolding of new forms of education and training,
significant changes in the status of the individuals and institutions involved in the education process,
and qualitative change in the lifestyle of individuals (Milana et al., 2018).
Lifelong learning and adult education have been simultaneously highly valued and strongly
contested. The processes of globalisation, rapid economic and technological change and an aging
workforce call for a permanent updating of knowledge and skills. Recent studies have found a
positive association between adult learning on the one hand and labour market outcomes, trust
and social justice on the other (Blossfeld et al., 2014; Boyadjieva & Ilieva-Trichkova, 2017; UNESCO
et al., 2015). However, there is also a criticism both of the epistemological status of the concept
and of the goals of European lifelong learning policies and practices. Whilst a lifelong learning
typology of regimes has been identified, which again includes a corporatist model controlled by the
social partners, and a market-based model based on competition (Verdier, 2013), lifelong learning
has been argued to be part of the neo-liberal project for which all that matters is the economy and
the market (Borg & Mayo, 2005). In this perspetcive, lifelong learning is a form of social control and
a mechanism promoting the marginalisation of the excluded and reasserting the social-reproductive
functions of education (Crowther, 2004; Jarvis, 2001).
Several studies have focused on identifying the factors at macro, meso and micro levels that affect
participation in different types of adult learning (Boeren, 2016). A reoccurring finding is that
educational advantages have cumulative character and advantages (or disadvantages) in early
adulthood continue to influence lifelong learning capability throughout life. Data also show that
there are big differences across countries in participation in lifelong learning. The practices of
lifelong learning are always produced by concrete historical circumstances related to the existing
links and interaction between specific national institutional systems, such as the educational
system, the labour market and social policies (Blossfeld et al., 2014; Rubenson & Desjardins, 2009).
In a world that is being transformed by the new digital technologies, the major challenge for policy-
makers and employers is to go beyond the instrumental and economised understanding of lifelong
learning, to acknowledge both its instrumental and substantial transformative/empowering value
at individual and societal level and to develop inclusive educational policies and practices.
5.3.4 The challenges ahead
Attempting to (re)configure skill systems to support the VET and workplace learning needs of the
digital economy is not easy. As the European Commission already implicitly recognises, whilst there
might be skill shortages in advanced digital skills in the future, the absolute number of jobs requiring
such skills will be very small, even if they extend from leading edge employers to cross into industries
not traditionally perceived as digitally-infused, for example farming and construction (Curtarelli and
Gualtieri, 2017). Moreover, given that it is difficult to predict what new jobs will emerge from the
digital economy, and in the context of an already widening spectrum of forms of work and
employment relationships, it will be consequently difficult, first, for the skills system to identify
future work’s skill types and levels and, second, for identifying responsibility within the ecosystem
for paying for and providing VET. As a consequence, matching supply with demand becomes even
more fraught than it is currently. The other challenge is that there are varying relationships across
EU countries between education, employment and work. In some countries the links are tight, in
other countries the relationship is loose (Bosch, 2017). Whether the links between education,
employment and work tighten or loosen as a consequence of digital transformation is an open
empirical question for the moment but will have consequences for the efficacy of skill ecosystems
within the digital economy. The (re) configuration of the systems needed to develop the skills of the
future also comes at a time when there are already pressures on national skill systems. Any potential
demand for new skills will have to compete against the Great Recession-induced squeeze on some
Member States’ public finances that has created a zero-sum competition between different
education and training stakeholders, most obviously schools, colleges, universities and even
employers, who have reduced their training budgets in some European countries. In addition, more
welfare spending will likely be directed to supporting the needs of an ageing population in Europe.
Thus some already ‘depleted welfare regimes’ will likely have extra demands make of them in
Europe (Keep, 2017, p.674; see also Gambin & Hogarth, 2017). Those Member States’ companies
and workforces that are already lagging behind in the digital transformation may fall further behind
a situation that the European Commission is keen to avoid EC, 2015).
Extending to the digital economy the general point made by Keep (2017) about skill systems,
policymaking will need to recognise that there are new as well as existing challenges and that there
may be no easy one-size fits all approach to future skills development through VET and workplace
learning. Any new skills settlement will be the outcome of accommodation between the competing
as well as sometimes overlapping interests of existing and new actors.
5.4. The creation and capture of value by companies
This section focuses on how value is currently created and captured and how BEYOND4.0 seeks to
develop alternative approaches to both. It starts by outlining current broad and narrow
understandings of value creation and capture before outlining BEYOND4.0s exploration of how
alternatives might be developed.
5.4.1 Understanding value creation and capture
The broad understanding of value creation and capture relates to various concepts of ecosystem:
business, entrepreneurial and innovation. All three types of ecosystem are closely interrelated yet
have subtle distinctive differences. As with the skill ecosystem concept, all three use and build on
the natural ecosystem analogy, defined as a biotic community, its physical environment, and all the
interactions possible in the complex of living and non-living components (Acs et al., 2017; Tansley,
1935). The use and application of the ecosystem analogy to business, entrepreneurial activity and
innovation started in the early 1990s and became popular in business and policy circles as from the
early 2000s onwards.
The concept of business ecosystems was first coined by Moore (1993) studying co-evolution in
social and economic systems, in particular networks of organisations that together constitute a
system of mutual support and co-evolving contributions. Moore saw the business ecosystem as a
form of organisation distinct from but parallel to markets and firms. More recent interpretations
frame business ecosystems as a form of economic coordination in which a firm’s ability to create
and appropriate value critically depends on different groups of actors that produce complementary
products or services (Acs et al., 2017; Iansiti & Levien, 2004). Following Adner (2017) business
ecosystems refer to the set of partners that need to be brought into alignment in order for a value
proposition to materialise in the marketplace.
The business ecosystem concept soon found its way into the innovation literature, with an
innovation ecosystem defined as a network of relationships through which information and talent
flow through systems of sustained value co-creation (Russell, et al., 2011, p.2). This network
comprises interacting, learning and innovating firms with stronger or weaker links between these
firms and science and technology institutions (such as universities and research labs) and with
technology firms. Within these systems, there is both a core and a wider setting. Some researchers
define systems of innovation narrowly as only comprising institutions explicitly involved in the
generation and diffusion of science and technology, with the bridge between research and industry
at the centre, and innovation viewed primarily as technological, whether product or process centred
(see Nelson, 1994). The broader approach and that adopted by BEYOND4.0understands
innovations more broadly: large or small; product, process or system; radical or incremental;
technological or organisational. All institutions and practices that affect the introduction or diffusion
of innovations are included, and the learning firms are placed at the core of the system (see
Freeman & Soete, 1997; Lundvall, 1992, 2007).
Closely related is the concept of entrepreneurial ecosystem, defined as a set of interdependent
actors and factors coordinated in such a way that they enable productive entrepreneurship within
a particular territory(Stam & Spigel, 2017, p.407). The entrepreneurial ecosystem concept shares
its focus on aggregate value creation within a particular region with the regional development and
regional ecosystem literature. More pronounced is its emphasis on the role of (individual)
entrepreneurs in creating value and a longer-term commitment to the region, with a less prominent
role for competition and value capture than in the business ecosystem concept (Acs, et al., 2017).
The narrow understanding of value creation and capture focuses on company-level practiceseven
if those practices provide demonstration effects for other companies in the same industry or come
to be systemwide for companies across industries. For BEYOND4.0 the obvious examples of the
former are Industrie 4.0 and Uberisation. Each in their own way offer examples of new business
models for value creation and capture. Business models are conceptual tools that express the
business logic of a specific firm and the way it operates to create value (Casadesus-Masanell &
Ricart, 2010; Pisano et al., 2015). As noted above, Industrie 4.0 offers companies an integrated
production system that, through the new digital technologies, links not only functions within
companies but also opens up these companies to suppliers and customers. Value can be created by
generating efficiency savings throughout the supply chain and by having direct links to customers
and being able to provide bespoke or customised goods and services. This digitisation of production
contrasts with companies that have hitherto used robotics and advanced automation for production
because those companies were ‘closedorganisations, with internal production only enveloped (see
Clark, 1995). Likewise noted above, with Uberisation, platforms are digital networks that coordinate
economic transactions usually matching the supply and demand of goods and services through
algorithms in which platforms companies claim to be brokers between supply and demand. In this
crowdsourcing of labour business model, through digital technology, a company externalises but
coordinates production of a good or services to previously unconnected, unorganised workers,
states Pisano et al. (2015). These authors claim that this model grows the value of ‘surplus’ (p.19)
that is, hitherto underused resources, citing the ‘share-riding’ of Uber as an example.
This value creation does however rest on the erosion of the standard employment relationship.
Relatedly, it is an approach to business that contrasts with the notions of the ‘classic’ firm, which
internalises resources and production to create vertically integrated enterprises. Importantly, this
internalisation reduces transaction costs and enables managerial improvements that thereby help
better value creation (Chandler, 1977; Coase, 1937).
In both casesIndustrie 4.0 and Uberisation, there is an expectation that these business models
will be adopted and diffused across companies in the manufacturing and services sectors.
Beyond digitalisation (but which is increasingly supported by diigitlaisation), there is also another
business model that is company-centric that is already pervasive and, some would argue, has been
hegemonic for the past 30 years, setting the benchmark for how value should be created and
captured: maximizing shareholder value (MSV) or, more prosaically, ‘financialisation’.
Financialisation focuses on increasing shareholder value, profits and flexibility, and is a form of value
creation and capture based on squeezing labour costs and revenues. It reinforces market discipline
and market attitudes within companies, whilst at the same time, promoting investors (shareholders)
as sovereign (Appelbaum, et al., 2013). As a business model it operates at three levels. At the macro
level value creation shifts way from human capital as its source. Instead, profits are derived from
financial assets driven by capital markets rather than product markets and production processes.
At the meso level of firm behaviour, financial engineering means that organisational restructuring
occurs trough delayering, disaggregation, mergers and acquisitions. At the micro level, the result is
work intensification, income and job insecurity for workers and, with pressures to turn quickly
generate profits, the squeezing of costs through redundancies and outsourcing (Cushen &
Thompson, 2016; see also Appelbaum, 2012). In requiring managers to ‘disgorge cash rather than
invest it’ (Jensen, 1986, p.323 quoted in Findlay, et al., 2017), and focus on value extraction rather
than value creation (Lazonick & Mazzucato, 2013), this business model overturns the operation
characteristic of the classic firm noted above. As a consequence, work and employment are
negatively impacted (Findlay et al., 2017). Moreover, Findlay et al. continue, there are wider impacts
that link work and welfare:
… many of these practices impose significant externalities on individuals on whom
organisational risk is heavily loaded; on families and communities disrupted by work
insecurity, unstable work patterns and low pay; and on the wider society, for example, in
necessitating welfare transfers to address low or variable pay, reducing tax revenue
opportunities, increasing health care demands and specifically health spend, as well as
limiting the return on public investment in education, learning and skills and in driving or
sustaining inequality, constraining growth at national level … (p.34)
Alternatives exist, for example stakeholder models. These models are based on collective interests
and social context in that they are sensitive not just to shareholders but the relationship between
firms and the wider institutional context. In this business model, all stakeholders have some type of
claim in relation to the operation and outcomes of firms. Advocacy of this model is often associated
with the ‘triple’ bottom line: economic, social and environmental. In this respect there is resonance
with the ecosystems approaches outlined above. However theory is currently undeveloped. There
is, as yet, no consensus on who constitutes the relevant stakeholders and the relationship between
these stakeholders and shareholders. In other words, ‘there is little specification in the literature
as to how these multiple objectives are weighed relative to each other’ according to Findlay et al.
(2017, p.13).
Two points follow. First, once again, choices can and are being made about the adoption of
particular business models and which have implications for value creation and capture. Second, that
theoretical development of alternatives to the current dominant model of maximising shareholder
value is required. This task is better enabled through the empirical evidence gathering of
5.4.2 A different route to value creation and capture
The ‘high road’ to inclusive growth requires more investment into the definition of what better
future jobs need. The discourse of Industrie 4.0 centres on an automated high road producing high-
valued-added goods and services. Emerging research by Hunt et al. (2019) on the impact of the
digitalisation of production in the private, public and voluntary sectors indicates that a key reason
that organisations introduce it do so to be able to develop new goods, services and processes (e.g.
predictive maintenance) or improve the quality of existing goods, services and processes and which
is more often than not realised. Moreover, whilst this digitalisation’s propensity to create and
destroy jobs is higher than for any other types of technology, with as many organisations reporting
job creation as job destruction (43% vs 40% respectively). Significantly, digitalisation had a positive
effect on some dimensions of job quality: increasing skills, and task complexity and control, and
improving work-life balance, pay and job security. Based on a single employer survey in the UK, how
pervasive these findings are across Europe is an open empirical question that BEYOND4.0 explores.
By contrast, Uberisation is a low roadstrategy, a model for producing cost-driven services. As
Wilson and Hogarth (2003) have pointed out, there is nothing wrong with some companies choosing
the low road.11 However, long-term difficulties arise for Europe as a whole if too many companies
choose this option. In the context of an over-educated workforce (Sarkar, 2017), and as noted
earlier, it represents a waste of human resources (OECD, 2011) and creates a ‘performance gap’ in
companies between what workers do and what they are capable of doing within different business
models (Livingstone, 2017). High road companies allow greater discretion from better skilled
employees (Milkman, 1998), which has morphed into ‘High Performance Work Systems’ (HPWS)
(Appelbaum et al., 2000). These HPWS are defined as an ‘approach to managing organisations that
aims to stimulate more effective employee involvement and commitment to achieve high levels of
performance’ (Belt & Giles, 2009: 17). Appelbaum et al. (2000) argue that HPWS provide the optimal
environment to elicit discretionary effort, underpinned by AMO. This AMO approach fits with the
workplace innovation movements and agendas that already exist at EU level and within Member
States (Oeij et al., 2017).
11 Osterman (2018) points out that there is a confusion in terminology. There are companies operating in sectors
with low wages and low profits. This does not mean that all companies follow ‘degrading’ policies for their
employees. It is necessary to identify if companies can follow ‘high road’ strategies in low pay sectors. The same
applies in high pay sectors: here companies can still follow ‘low road’ strategies. In such an approach. ‘low’ and ‘high
road’ are more strategies that reflect the preparedness of companies to look at their organisational practices and
the way income is redistributed.
The EU can decide to block off the low road through legislation and/or pave the high road by
encouraging environments that generate better jobs (cf. Carré et al., 2013). The introduction of the
new Directive on is one attempt to block off the low road and, in part, aimed at the worst examples
of the gig economy. A key task of BEYOND4.0 is to develop evidence-based recommendations as to
how to pave the high road.
BEYOND4.0’s focal point for value creation is the entrepreneurial and the business ecosystem. Such
value creation depends on productive solutions that shape innovation within and between
companies and other stakeholders. Decisions though occur at different levels. BEYOND4.0 draws
on three frameworks for this investigation: the innovation diffusion framework to structure findings
into an evolutionary maturity model of Industrie 4.0 implementation; a more radical model of
winner-takes-all, usually led an outsider challenger (start-tup or scale-up), and a dynamic resource-
based framework detailing the organisational capabilities underlying different Industrie 4.0 based
strategies and maturity stages. These perspectives help understanding of developments at
ecosystem and company levels.
Applying this approach to Uberisation is tricky. However BEYOND4.0 posits that existing platforms
economy business models can be illusory. Often in discourse about innovation and the future of
work, it is taken for granted that it is the private sector entrepreneur, owner, or shareholder who is
taking on the risk while other actors (the state, taxpayers, workers, users/customers) are relatively
passive, merely enjoying the consumption ‘spill-over’ benefits from ‘open innovation’ and new
technology but contributing little to the innovation itself. This narrative is often used to justify the
sometimes super-normal profits and/or share values that companies are able to generate from new
forms of ‘intangible assets’ (Haskel & Westlake, 2018), not least in the platform economy.
BEYOND4.0 takes a critical approach to understanding the processes of value generation and
extraction in the digital industries, recognising that innovation in digital/internet industries is
characterised by strong network effects (the more subscribers, the more value of the product) and
first-mover advantages. Anyone who gains an initial advantage, in setting a standard or capturing
part of a ‘sticky’ market, can be very hard to displace (Mazzucato, 2018). As a platform economy
firm’s market share rises, so does its capacity to attract users, which in turn increases its market
dominance. Many platform economy business models rely on loss-leader models funded by massive
venture capital or other leveraged forms of investment to establish monopolistic control over
markets that then enables the rapid extraction of value via returns to scale (Langley & Leyshon,
2017). This possibility is partially due to the way debt is favoured over equity in modern taxation
regimes, including the EU.12 It is not clear that such business models necessarily generate optimal
product quality, enable competitive markets or are likely to support high quality, secure
employment. Indeed, given the need for fast returns to investors and creditors, let alone create
high quality jobs. Many platform economy business models operate in ‘multi-sided’ markets,
developing both the supply and demand side as intermediaries/brokers of information. The service
they provide is ostensibly free, promoted as part of a participatory culture in the putative ‘sharing
economy’. However, these firms rely upon users voluntarily providing commercially valuable
information about themselves or their (consumer) preferences in return for access to the platform.
As such, users could be considered to be key but unrewarded, stakeholders in the creation of value
for these platforms. Paradoxically, in national accounting terms, platform economy firms’ value is
12 However see reference to a new perspective in Saez & Zucman (2019).
recorded as relating to profits made from selling advertising and information, not activities
generating added value or productivity increases for the wider economy, rather than from the direct
services they provide (which are free) (Mazzucato, 2018).
Although the venture capitalist business model allows rapid market capitalisation, it is not a given.
Alternative business models exist. For example, a digital taxi hailing app can be owned by: a US start-
up owned and funded by Silicon Valley venture capitalists or a metropolitan municipal government
or a city-region cooperative of taxi drivers. Indeed, Schor and Fitzmaurice (2015) have been arguing
for a (re-)emergence within the platform economy of more genuine collaboration and peer-to-peer
sharing, i.e. a new high road appraoch. These different ownership, financing and governance
structures may lead to different business models and impacts on job quality (Warhurst et al., 2017)
and to other non-commercial considerations such as environmental sustainability. The business
models need to be identified that address the value creation/value capture issue (or the problem
of economic rents), with public policy that disincentivises unsustainable business models (Findlay
et al., 2017). As Soros (2018) points out, platform companies are innovative and transformative of
business in the short-term. However, because of their monopolistic behaviour, they are also a
potential obstacle to long-term innovation. Their business model is unsustainable. These companies
have neither the will nor the inclination to deliver wider socio-economic benefits and it falls to
government policy, particularly EU, to ensure that technological innovations do so.
BEYOND4.0 explores the possibilities for public policy to shape and co-create the market
development of platform economies along different models reflecting wider socio-economic
objectives. This exploration will include an examination of how such companies should be effectively
taxed and regulated in regard to issues as privacy, information sharing and intellectual property.
One option for example, could be a ‘harvest tax’ in which part of these companies’ captured wealth
is redistributed to the state on the basis on the number of users and their data that are harvested
by these companies. Another idea developed by Saez and Zucman (2019) is that rather than trying
to tax companies that can evade taxes by allocating financial losses from all over the world to their
accounts in other countries, countries can calculate the ‘owed taxes’ on a world scale, and then just
tax these companies for this rate. For example if IKEA earns 20 per cent of its income in the US, the
American IRS should just claim 20 per cent of the tax rebate from IKEA. This approach is easy to do
because all the tax data are available, no new legislation is needed, it is completely legal and, above
all, it is lucrative for a state to do. It also reduces the drive for these companies to develop tax
constructions etc. and so the need for tax evasion disappears (Witteman, 2019).
More generally, BEYOND4.0 challenges standard narratives about the role of the state and wider
public sector in the innovation and value creation process in relation to technological and digital
transformations. The ambition to achieve a particular type of economic growth (smart, inclusive,
sustainable) is a direct admission that economic growth has not only a rate but also a direction. In
this context, industrial and innovation strategies are key pillars to achieve desirable change. In
particular, by identifying and articulating new missions that galvanise production, distribution and
consumption patterns across various sectors (Mazzucato & Perez, 2014). Making this shift requires
investment by both private and public sector actors. Many successful tech companies, such as
Apple, and technological innovations more generally (e.g. the internet) have benefitted from early
stage public investment in high-risk R&D that the private sector is generally unwilling to provide
(Mazzucato & Perez, 2014). The public sector can thus be seen to have a role in ‘market shaping’
just as much as its traditional role in solving ‘market failures’ or providing a safety net for citizens in
the face of rising inequality. Recognition of this role is important when assessing risks and rewards
of different actors in an innovation process being a collective not individual process (Lazonick &
Mazzucato, 2013). What is happening at the ecosystem level depends on understanding this
institutional surrounding.
6. Conclusion
BEYOND4.0 examines the impact of new digital technologies on the future of jobs, business models
and welfare in the EU. It is ambitious in its scope and multi-faceted in its approach. This guidance
paper has provided outlines of the key concepts, issues and developments that are the starting
point for the project. Whilst digitalisation impacts on many facets of work and life, the particular
foci of BEYOND4.0 is the digitalisation of production, commonly termed Industrie 4.0, and the
digitalisation of work, now not uncommonly termed Uberisation. These foci are selected because
of their prominence in current policy and academic debate and because, within those debates, they
also cause much consternation: in different ways, there are concerns that both developments
potentially herald the death of jobs and the hollowing out of welfare in Europe and elsewhere.
Nevertheless, while concerns are high, evidence is short. BEYOND4.0 addresses this evidence deficit.
Taking a pan-European perspective, it has four specific areas of empirical enquiry: the quality,
content, and distribution of work; the skill needs of the labour market; the education and training
to support development of these skills; and the creation and capture (extraction) of value (and
wealth) by companies. Moreover its analysis is rooted in historical context and the lessons from
previous technological revolutions. The evidence generated will significantly advance scientific
understanding of digitalisation and its impact on work and welfare. The evidence will also help
identify other, non-dystopian futures for Europe from that in which work and welfare collapse.
In this regard, whist it primary function is to generate new scientific understanding about the future
of work and welfare beyond current debates about Industrie 4.0 and Uberisation, BEYOND4.0 is also
explicitly intends to shape EU policy, specifically the delivery of an inclusive European future for all
as the digital transformation unfolds. By generating new scientific evidence and policy development
around these issues, BEYOND4.0 helps further the EC’s Europe 2020 strategy promoting smart,
sustainable and inclusive growth by responding to the challenges and maximising the opportunities
of digitalisation in Europe for the next decade and beyond.
The authors would like to thank Peter Fairbrother of the Royal Melbourne Institute of Technology,
Angie Knox of Sydney University, and Wil Hunt and Sudipa Sarkar both of Warwick University for
discussions in and around the preparation of this Working Paper.
Acemoglu, D. & Autor, D. (2010). Skills, Tasks and Technologies: Implications for employment and
earnings. NBER Working Paper 16082. Cambridge, MA: National Bureau of Economic Research.
Acs, Z.J., Stam, E., Audretsch, D.B., & O’Connor, A. (2017). The lineages of the entrepreneurial
ecosystem approach. Small Bus Econ, 49, 110.
Adner, R. (2017). Ecosystem as structure: an actionable construct for strategy. Journal of
Management, 43(1): 39-58.
Alli, B.O. (2008). Fundamental principles of occupational health and safety. Geneva: International
Labour Office, XX+200 p.
Anderson, P. & Warhurst, C. (2012). ‘Lost in Translation: Skills policy and the shift to skill
ecosystems. In T. Dolphin & D. Nash. (eds). Complex New World: Translating new economic thinking
into public policy, London: IPPR.
Annunziata, M. & Bourgeois, H. (2018). The future of work: how G20 countries can leverage digital-
industrial innovatins into stronger high-quality jobs growth. Economics, 12:1-23.
Appelbaum, E. (2012). Reducing Inequality and Insecurity: Rethinking Labour and Employment
Policy for the 21st Century. Work and Occupations, 39(4): 311320.
Appelbaum, E., Bailey, T., Berg, P. & Kalleberg, A.L. (2000). Manufacturing advantage. Ithaca: Cornell
University Press.
Appelbaum, E., Batt, R. & Clark, I. (2013). Implications of financial capitalism for employment
relations research: evidence from breach of trust and implicit contracts in private equity buyouts.
British Journal of Industrial Relations, 51(3): 498-518.
Ashton, D., Lloyd, C. & Warhurst, C. (2017). Business Strategies and Skills. In C. Warhurst, K.
Mayhew, D. Finegold & J. Buchanan. (eds). Oxford Handbook of Skills and Training, Oxford: Oxford
University Press.
Autor, D. (2010). The Polarization of Job Opportunities in the U.S. Labour Market Implications for
Employment and Earnings. Washinton, D.C.: MIT Department of Economics and National Bureau of
Economic Research, The Center for American Progress and The Hamilton Project.
Autor, D. (2013). The “task approach” to labour markets: an overview. Journal of Labour Market
Research, 46(3): 185-199.
Autor, D.H. (2015). ‘Why are there still so many jobs? The history and future of workplace
automation.’ Journal of Economic Perspectives, 29(3): 3-30.
Avent, R. (2014). The third great wave. The Economist, 2 October 2014, p. 4.
Badham, R.J. (2006). Technology and the transformation of work. In S. Ackroyd, R. Batt, P.
Thompson & P. Tolbert (eds.) The Oxford handbook of Work & Organisation. Oxford: Oxford
University Press.
Barnes, S-A., Green, A. & de Hoyos, M. (2015). Crowdsourcing and work: individual factors and
circumstances influencing employability. New Technology, Work and Employment 30(1): 16-31.
Behle, H. (2016). LEGACY: Employability of higher education graduates. Working Paper, Institute
for Employment Research, University of Warwick.
Bernhardt, A. (2016). It’s not all about Uber. Perspectives on Work, 20: 14-17 & 76.
Berting, J. (1993). Organization studies and the ideology of technological determinism. In S.M.
Lindenberg & H. Schreuder. (eds). Interdisciplinary perspectives on organization studies. Oxford:
Pergamon Press.
Beuker, L., Franssen, M., Kirov, V. & Naedenoen, F. (2019). Digitalisation and Restructuring: Which
social dialogue? Work package 1: Transnational analysis. Synthesis report. DIRESOC: Sofia. Available
(accessed 2 September 2019).
Bijker, W.E., Hughes, T.P. & Pinch, T.J. (2012). The Social Construction of Technological Systems: New
Directions in the Sociology and History of Technology. Massachusetts; MIT Press.
Beynon, H. (1973). Working for Ford. Harmondsworth: Pelican.
Blanden, J., A. Goodman, P. Gregg & S. Machin (2005). ‘Changes in intergenerational income
mobility in Britain’ in M. Corak (ed.) Generational income mobility in North America and Europe,
Cambridge, MA: Cambridge University Press.
Blau, F.D., Kahn, L.M. (2017) The Gender Wage Gap: Extent, Trends, and Explanations. Journal of
Economic Literature, 55 (3): 789-865. DOI: 10.1257/jel.20160995
Bloom, N., Brynjolfsson, E., Foster, L., Jarmin, R., Patnaik, M., Saporta-Eksten, I. & Van Reenen, J.
(2019). What Drives Differences in Management Practices? American Economic Review 109(5):
Blossfeld, H.P., Kilpi-Jakonen, E., Vono de Vilhena, D., & Buchholz, S. (2014). Adult learning in
modern societies. An international comparison from a life-course perspective. Cheltenham: Edward
Boeren, E. (2016). Lifelong learning participation in a changing policy context: an interdisciplinary
theory. London: Palgrave-Macmillan.
Bonoli, G., George, V. & Taylor-Gooby, P. (2000). European Welfare Futures: Towards a theory of
retrenchment, Cambridge: Polity.
Borg, C. & Mayo, P. (2005). The EU memorandum on lifelong learning. Old wine in new bottles?
Globalisation. Societies and Education, 3(2): 203225.
Bosch, G. (2017). Different National Skill Systems. In C. Warhurst, K. Mayhew, D. Finegold & J.
Buchanan. (eds). Oxford Handbook of Skills and Training. Oxford: Oxford University Press.
Bosworth, D., Sarkar, S. and Warhurst, C. (2019) Making the business case: good work and
productivity at the sector level in the UK. A draft summary of initial findings, Dunfermline/London:
Carnegie UK Trust-RSA.
Boyadjieva, P. & Ilieva-Trichkova, P. (2017). Between inclusion and fairness: Social justice
perspective to participation in adult education. Adult Education Quarterly, 67(2): 97117.
Brynjolfsson, E., & Milgrom, P. (2013). Complementarity in organizations’. The Handbook of
Organizational Economics, 11-55.
Brynjolfsson, E. & McAfee, A. (2014). The Second Machine Age. New York: W.W. Norton & Co.
Bundesanstalt für Arbeitsschutz und Arbeitsmedizin. (2010). Ratgeber zur Gefährdungsbeurteilung.
Handbuch für Arbeitsschutzfachleute (Guideline risk assessment. Manual for OSH professionals).
Bremerhaven: Wirtschaftsverlag NW Verlag für neue Wissenschaft GmbH. Available from: (accessed
15 February 2019).
Burchell, B, Sehnbruch, K, Piasna, A & Agloni, N (2014). The quality of employment and decent
work: Definitions, methodologies and ongoing debates. Cambridge Journal of Economics, 38(2):
Canivet, C., Choi, B., Karasek, R., Moghaddassi, M., Stalans-Nyman, C. & Ostergren P-O. (2013). Can
high psychological demands, low decision latitude and high job strain predict disability pensions? A
12-year follow-up. Int. Arch. Occup. Environm. Health; 86(3): 307-319.
Cantillon, B. & Vandenbroucke, F. (eds). (2014). Reconciling Work and Poverty Reduction. How
Successful are European Welfare States? Oxford: Oxford University Press.
Cappelli, P. & Keller, J. R. (2013). Classifying work in the new economy. Academy of Management
Review, 38(4): 575-596.
Carré, F., Findlay, P., Tilly, C. & Warhurst, C. (2013) ‘Job quality: scenarios, analysis and interventions’
in C. Warhurst, P. Findlay, C. Tilly & F. Carré (eds) Are bad jobs inevitable? London: Palgrave.
Casadesus-Masanell, R. & Ricart, J.E. (2010). From strategy to business models and onto tactics.
Long Range Planning, 43(2-3): 195-215.
Castles, F., Leibfried, S., Lewis, J. Obinger, H. & Pierson, C. (eds). (2010). The Oxford Handbook of
the Welfare State. Oxford: Oxford University Press.
Cedefop. (2017). Skills panorama glossary,
Chandler, A. (1977). The Visible Hand: The Managerial Revolution in American Business, Cambridge,
Mass: Harvard University Press.
Chartered Institute for Personnel & Development (CIPD) (2018). UK Working Lives : In search of job
quality, London : CIPD.
Cherns, A. (1976). The principles of sociotechnical design. Human Relations, 29(8): 783-792.
Child, J. (1972). Organisational Structure, Environment and Performance: The role of strategic
choice. Sociology, 6(1): 1-22.
Clark, J. (1995). Managing Innovation and Change: People, Technology and Strategy. London: Sage.
Clarke, S. (2010). An integrative model of safety climate: linking psychological climate and work
attitudes to individual safety outcomes using meta-analysis. Journal of Occupational and
Organizational Psychology, 83: 553578.
Coase, R.H. (1937). The nature of the firm. Economica, 4(16): 386-405.
Coomber, B., & Barriball, K.L. (2007). ‘Impact of job satisfaction components on intent to leave and
turnover for hospital-based nurses: A review of the research literature. International Journal of
Nursing Studies, 44: 297314.
Crouch, C. (2006). ‘Skill formation systems. In S. Ackroyd, R. Batt, P. Thompson & P. Tolbert (eds.)
The Oxford handbook of Work & Organisation. Oxford: Oxford University Press.
Crowther, J. (2004). ‘‘In and against’ Lifelong learning: flexibility and corrosion of character’.
International Journal of Lifelong Education, 23(2): 125136.
Curtarelli, M. & Gualtieri, V. with Jannati, M. S. & Donlevy, V. (2017). ICT for work: Digital skills in
the workplace. Final Report from study prepared for the Director General for Communications
Networks, Content and Technology (DG CONNECT), European Commission. Luxembourg:
Publications Office of the European Union.
Cushen, J. & Thompson, P. (2016). Financialisation and Value: Why labour and the labour process
still matter. Work, Employment and Society, 30(2): 352-365.
Davies, R. (2015). Industry 4.0 Digitalisation for productivity and growth, Briefing, European
Parliamentary Research Service.
De Gieter, S., Hofmans, J. & Pepermans, R. (2011). Revisiting the impact of job satisfaction and
organizational commitment on nurse turnover intention: An individual differences analysis.
International Journal of Nursing Studies, 48: 15621569.
Degryse, C. (2016). Digitalisation of the economy and its impact on labour markets (ETUI working
paper). Brussels: ETUI.
Deloitte. (2014). Agiletown: The relentless march of technology and London’s response. London:
London Futures.
Dhondt, S. (2019). ‘Possibilities to collect information about platforms in Europe: Learning from the
US.’ Research note 2019/0001. Leiden: TNO.
Dhondt, S. & Van Hootegem, G. (2015). Reshaping workplaces: workplace innovation as designed
by scientists and practitioners. European Journal of Workplace Innovation, 1(1): 17-24.
Dregger, J., Niehaus, J., Ittermann, P., Hirsch-Kreinsen, H., & ten Hompel, M. (2016). The Digitization
of manufacturing and its societal challenges. A framework for the future of industrial labour.
Vancouver: IEEE.
Duguet, E., & Greenan, N. (1997). ‘Le biais technologique: une analyse économétrique sur données
individuelles’. Revue économique, 1061-1089
Duijts, S.F.A., Kant, I.J., Swaen, G.M.H., Brandt, P.A. van den & Zeegers, P.A. (2007). A meta-analysis
of observational studies identifies predictors of sickness absence. J Clinical Epidemiology, 60: 1105-
Dunlop, T. (2016). Why the Future is Workless. Sydney: University of New South Wales Press.
Eisenbrey, R. & Mishel, L. (2016). Uber business model does not justify a new ‘independent worker
category. Economic Policy Institute.
Elias, P. & Day, R. (2017). Defining occupations in science, technology, engineering, mathematics,
medicine and health via the Standard Occupational Classification. Institute for Employment
Research, University of Warwick.
Ennen, E., & Richter, A. (2010). ‘The whole is more than the sum of its partsor is it? A review of
the empirical literature on complementarities in organizations’. Journal of Management, 36(1), 207-
Erhel, C. & Guergoat-Larivière, M. (2010). ‘Job quality and labour market performance. CEPS
Working Document No 330, Centre for European Policy Studies, Paris.
Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Princeton: Princeton University
Esping-Andersen, G. (2009). The Incomplete Revolution: Adapting welfare states to women’s new
roles. Cambridge: Polity Press.
Esping-Andersen, G. with Gallie, D., Hemerijck, A. & Myles, J. (2002). Why we need a new welfare
state. Oxford: Oxford University Press.
Eurofound. (2012). Trends in job quality in Europe. Dublin: European Foundation for the
Improvement of Living and Working Conditions.
Eurofound. (2016). What do Europeans do at work? A task-based analysis: European Jobs Monitor
2016. Luxembourg: Publications Office of the European Union.
Eurofound. (2018a). Does employment staus matter for job quality? Luxemburg: Publications Office
of the European Union.
Eurofound. (2018b). Automation, digitalisation and platforms: Implications for work and
employment. Luxembourg: Publications Office of the European Union.
Eurofound. (2018c). Upward convergence in the EU: Concepts, measurements and indicators.
Dublin: European Foundation for the Improvement of Living and Working Conditions.
European Commission. (EC) (2009). Causes and circumstances of accidents at work in the EU,
Directorate-Greneral for Employment, Social Affairs and Equal Opportunities F4 unit. Luxembourg:
Office for Official Publications of the European Communities. Available from:
European Commission. (EC) (2010). Europe 2020: A strategy for smart, sustainable and inclusive
growth. Brussels: Publications Office of the European Union.
European Commission. (2015). A Digital Single Market Strategy for Europe: Analysis and Evidence.
Commission Staff Working Document SWD(2015)100. Luxembourg: Publications Office of the
European Union.
European Commission. (EC) (2019a). The Impact of the Digital Transformation on EU Labour
Markets, Report of the High-Level Expert Group on The impact of the Digital Transformation on EU
Labour Markets, Luxembourg: Publications Office of the European Union.
European Commission. (EC) (2019b). Digital Agenda for Europe. Factsheets of the European Union.
Brussels: Publications Office of the European Union (accessed 4 February 2019).
European Commission. (n.d.). Regional Ecosystem Scoreboard. Brussels: European Commission:
Available from:
ecosystem-scoreboard_en (accessed 4 February 2019).
European Commission. (2015a). A Digital Single Market Strategy for Europe. Communication from
the Commission to the European Parliament, the Council, the European Economic and Social
Committee and the Committee of the Regions. COM/2015/0192 final. Luxembourg: Publications
Office of the European Union.
European Commission. (2015b). A Digital Single Market Strategy for Europe: Analysis and Evidence.
Commission Staff Working Document SWD (2015)100. Luxembourg: Publications Office of the
European Union.
European Economic and Social Committee. (2011). Opinion of the European Economic and Social
Committee on Innovative workplace as a source of productivity and quality jobs, SC/034 CESE 543.
Official Journal of the European Journal C132/22.
European Political Strategy Centre. (EPSC) (2016). The Future of Work. EPSC Briefing Notes,
European Commission.
Eurostat. (2004). Statistical analysis of socio-economic costs of accidents at work in the European
Union. Luxembourg: Office for Official Publications of the European Union.
Eurostat. (2010). Health and safety at work in Europe (1999-2007) A statistical portrait.
Luxembourg: Office for Official Publications of the European Communities. Available from: (accessed 15
February 2019).
Felstead, A., Gallie, D. & Green, F. (2017). Measuring Skills Stock, Job Skills and Skills Mismatch. In
C. Warhurst, K. Mayhew, D. Finegold & J. Buchanan. (eds.). Oxford Handbook of Skills and Training,
Oxford: Oxford University Press.
Felstiner, A. (2011). Working the crowd: Employment and labour law in the crowdsourcing
industry. Berkeley Journal of Employment & Labour Law, 32(1): 143203.
Fernandez-Macías, E., Bisello, M., Sarkar, S. & Torrejón, S. (2016). Methodology of the construction
of task indices for the European Jobs Monitor. Luxembourg: Publications Office of the European
Findlay, P., Kalleberg, A. & Warhurst, C. (2013). The challenge of job quality. Human Relations,
66(4): 441-451.
Findlay, P., Thompson, P., Coper, C. & Pascoe-Deslauriers, R. (2017). Creating and capturing value
at work: who benefits? London: CIPD.
Fiore-Gartland, B. & Tanweer, A. (2018). Community-level data science and its spheres of influence:
beyond novelty squared. eScience Institute. Available from:
beyond-novelty-squared/ (accessed 31 December 2018).
Forde, C., Stuart, M., Joyce, S. Oliver, L., Valizade, D. Alberti, G., Hardy, K., Trappmann, V., Umney,
C. & Carson, C. with Katja, J. & Yordanova, G. (2017). The social protection of workers in the platform
economy, Policy Department A: Economic and Social Policy, European Parliament, Brussels.
Freeman, C. & Soete, L. (1997). The Economics of Industrial Innovation [Third Edition]. London:
Frey, C.B. & Osborne, M.A. (2013). The Future of Employment: How susceptible are jobs to
computerisation. Oxford Martin Programme on the Impacts of Future Technology Working Paper.
Oxford Martin Programme on Technology and Employment, University of Oxford.
Gambin, L. & Hogarth, T. (2017). Who Pays for Skills? Differing perspectives on who should pay and
why. In C. Warhurst, K. Mayhew, D. Finegold & J. Buchanan. (eds). Oxford Handbook of Skills and
Training, Oxford: Oxford University Press.
Gambin, L., Hogarth, T., Murphy, L. Spreadbury, Warhurst, C. & Winterbotham, M. (2016). Research
to understand the extent, nature and impact of skills mismatches in the economy, BIS Research
Paper no.265, London: Department for Business Innovation & Skills.
Goldthorpe, J. & Mills, C. (2008) ‘Trends in Intergenerational Class Mobility in Modern Britain:
Evidence From National Surveys, 19722005’, National Institute Economic Review, 205(1): 83-100.
Goos, M., Manning, A. & Salomons, A. (2009). The Polarization of the European Labor Market.
American Economic Review, 99(2): 58-63.
Goos, M., Manning, A. & Salomons, A. (2014). Explaining Job Polarization: Routine-Biased
Technological Change and Offshoring. American Economic Review, 104 (8): 2509-2526.
Gordon, R. (2012). Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six
Headwinds, NBER Working Paper No. 18315Cambridge, MA..
Green, F. (2011). ‘What is skill? An interdisciplinary synthesis’, LLAKES Research Paper 20, Institute
of Education, University of London.
Green, F., Mostafa, T., Parent-Thirion, A., Vermeylen, G., Van Houten, G., Biletta, I. & Lyly-
Yrjanainen, M. (2013). Is job quality becoming more unequal?. ILR Review, 66(4): 753-784.
Greenan N. (2003), Organizational Change, Technology, Employment and Skills: an Empirical Study
of French Manufacturing, Cambridge Journal of Economics, Vol. 27, N°2, pp.287-316.
Greenan, N., & Seghir, M. (2017). Measuring vulnerability to adverse working conditions: evidence
from European countries. Document de travail du CEET No. 193, Paris : CEET.
Greenan, N., Kalugina, E. & Walkowiak, E. (2013). Has the quality of working life improved in the
EU-15 between 1995 and 2005?Industrial and Corporate Change, 23(2): 399-428.
Greve, B. (ed.). (2018). Routledge Handbook of the Welfare State (second edition). Oxford:
Griliches, Z. (1969). ‘Capital-skill complementarity’. The Review of Economics and Statistics, 51(4):
Grote, G. & Guest, D. (2017). ‘The case for reinvigorating quality of working life research. Human
Relations, 70(2): 149-167.
Grugulis, I., Warhurst, C. & Keep, E. (2004). ‘What’s Happening to Skill?’ In C. Warhurst, E. Keep & I.
Grugulis. (eds). The Skills That Matter, London: Palgrave.
Guiochet, J. (2016) Hazard analysis of human-robot interactions with HAZOP-UML, Safety Science,
Halford, S., Hudson, M., Leonard, P., Parry, J. & Taylor, R. (2016). The New Dynamics of Work: A
Scoping Study. Work Futures Research Centre Working Paper.
Hamermesh, D. (1993). Labor demand, Princeton, NJ: Princeton University Press.
Harms-Ringdahl, L. (2001). Safety Analysis: Principles and Practice In Occupational Safety. London:
Taylor & Francis.
Harris, S.D. & Krueger, A.B. (2015). ‘Proposal for modernising labour laws for twenty-first-century
work: the “independent worker. Discussion paper 2015-10, Washington, DC: Brookings Institution.
Haskel, J., & Westlake, S. (2018). Capitalism Without Capital: The Rise of the Intangible Economy.
Princeton: Princeton University Press.
Hemerijck, A. (2013). Changing Welfare States. Oxford: Oxford University Press.
Hermann, M., Pentek, T., Otto, B. (2016). Design Principles for Industrie 4.0 Scenarios. Proceedings
of the Annual Hawaii International Conference on System Sciences, 7427673, pp. 3928-3937 DOI
Hirsch-Kreinsen, H. (2016). Digitization of industrial work: development paths and prospects.
Journal for Labour Market Research, 49: 1-14.
Hirsch-Kreinsen, H. (2016). Digitization of industrial work: development paths and prospects.
Journal for Labour Market Research, 49(1): 1-14.
Holm, J. R., & Lorenz, E. (2015). Has “Discretionary Learning” declined during the Lisbon Agenda?
A cross-sectional and longitudinal study of work organization in European nations. Industrial and
Corporate Change, 24(6): 1179-1214.
Holman, D. (2013). Job types and job quality in Europe. Human Relations, 66(4): 475-502.
Howaldt, J. & Schwarz, M. (2010). Social Innovation: Concepts, Research Fields and International
Trends (Studies for Innovation in a Modern Working Environment International Monitoring Vol.
5). Dortmund. Available from: http://www.sfs.tu-
(accessed on 15 February 2019).
Hunt, W., Warhurst, C. & Sarkar, S. (2018). Innovation regime and vulnerable workers’ labour
market inclusion and job quality, QuInnE Working Paper No. 13,
Hunt, W., Warhurst, C. & Sarkar, S. (2019). People and Machines: from hype to reality: Technical
Report. London: CIPD.
Huws, U., Spencer, N. H. & Joyce, S. (2016). Crowd work in Europe: Preliminary results from a survey
in the UK, Sweden, Germany, Austria and the Netherlands. Brussels: European Foundation for
Progressive Studies.
Iansiti, M., Levien, R. (2004). Strategy as ecology. Harvard Business Review, 82:3, March, 1-11.
International Labour Organization. (ILO) (2016). Non-standard employment around the world,
Understanding challenges, shaping prospects. Geneva: International Labour Office.)
Ittermann, P., Niehaus, J., Hirsch-Kreinsen, H., Dregger, J. & ten Hompel, M. (2016). Social
Manufacturing and Logistics. Gestaltung von Arbeit in der digitalen Produktion und Logistik.
Soziologisches Arbeitspapier, Nr. 47, Dortmund.
Jacobs, M. & Mazzucato, M. (2016). 'Rethinking Capitalism: Economics and Policy for Sustainable
and Inclusive Growth. Chichester: Wiley & Sons.
James, S. Warhurst, C., Tholen, G. & Commander, J. (2013). ‘What We Know and What We Need to
Know About Graduate Skills. Work, Employment and Society, 27(6): 952-963.
Jarvis, P. (2001). The changing educational scene. In P. Jarvis. (ed.). The age of learning: Education
and the knowledge society (pp. 2738). London, Routledge.
Jenkins, C. & Sherman, B. (1979). The Collapse of Work. London: Eyre Methuen.
Kalleberg, A.L. (2009). Precarious work, insecure workers: employment relations in transition.
American Sociological Review, 74(1): 1-22.
Kaptelinin, V., & Nardi, B. A. (2006). Acting with technology: Activity theory and interaction design.
Cambridge, MA: MIT Press.
Katz, R., Koutroumpis, P. & Callorda, F. (2014). Using a digitization index to measure the economic
and social impact of digital agendas. Info, 16(1): 3244.
Keep, E. (2017). In C. Warhurst, K. Mayhew, D. Finegold & J. Buchanan. (eds) Oxford Handbook of
Skills and Training, Oxford: Oxford University Press.
Kerr, C. Dunlop, J.T., Harbison, F. & Myers, C.A. (1960). Industrial and Industrial Man.
Harmondsworth: Pelican.
Kivimaki, M., Nyberg, S.T., Batty, G.D., et al.,(2012). Job strain as a risk factor for future coronary
heart disease: collabourative meta-analysis of 2358 events in 197,473 men and women. The Lancet,
380: 1491-97.
Knox, A., Warhurst, C., Nickson, D. & Dutton, E. (2015). More than a feeling: using hotel room
attendants to improve understanding of job quality. International Journal of Human Resource
Management, 26(12): 1547-1567.
Kondratiev, N. (1935). The Long Waves in Economic Life. Review of Economic Statistics, 17: 105-15.
Kopp, R., Dhondt, S., Hirsch-Kreinsen, H., Kohlgrüber, M., Preenen, P. (2019). ‘Sociotechnical
perspectives on digitalisation and industry 4.0’. International Journal of Technology Transfer and
Commercialisation, 16(3): 290-309.
Kuipers, H., Van Amelsvoort, P. & Kramer, E.-H. (2018). Het nieuwe organiseren. Leuven: Acco.
Lafer, G. (2004). ‘What is “skill’? Training for discipline in the low-wage labour market. In C.
Warhurst, E. Keep & I. Grugulis. (eds). The Skills That Matter, London: Palgrave.
Lamb, M.E. (2012). Mothers, Fathers, Families, and Circumstances: Factors Affecting Children's
Adjustment. Applied Developmental Science 16:2, April, 98-111
Langley, P., Leyshon, A. (2017) Platform capitalism: the intermediation and capitalisation of digital
economic circulation. Finance and society, 3:1, 11-31.
Lazonick, W. & Mazzucato, M. (2013). The risk-reward nexus in the innovation-inequality
relationship: who takes the risks? Who gets the rewards?Industrial and Corporate Change, 22(4):
Lever, R. (2017). Tech world debate on robots and jobs heats up. Phys.Org. 26 March.
Leveson, N.G. (2011) Engineering a Safer World: Systems Thinking Applied to Safety Cambridge
(MA): The MIT Press
Levy, F. & Murnane R. J. (1992). U.S. Earnings and Earnings Inequality: A review of recent trends
and proposed explanations. Journal of Economic Literature, 30(3): 1333-1381.
Levy, F. & Murnane, R. J. (2004). The New Division of Labour. New York: Russell Sage.
Levy, F., & Murnane, R. J. (2013). Dancing with robots: Human skills for computerized work.
Cambridge, MA: Third Way.
Livingstone, D.W. (2017) ‘Skill Under-utilization’ in C. Warhurst, K. Mayhew, D. Finegold & J.
Buchanan (eds) Oxford Handbook of Skills and Training, Oxford: Oxford University Press.
Lundvall, B-A. (1992). National Systems of Innovation: Toward a Theory of Innovation and Interactive
Learning. London: Pinter.
Lundvall, B-A. (2007). National Innovation Systems-Analytical Concept and Development Tool.
Industry and Innovation, 14(1): 95119.
Lundvall, B-A., Rasmussen, P. & Lorenz, P. (2008). Education in the learning economy: a European
perspective. Policy Futures in Education, 6(6): 681-700.
Madsen, I.E.H., Nyberg, S.T., Magnusson Hanson, L.L., Ahola, K., et al.,(2017). Job strain as a risk
factor for clinical depression: systematic review and meta-analysis with additional individual
participant data. Psychological Medicine, 47(8): 1342-1356.
Manyika, J., Chui, M., Miremadi, M., Bughin, J., George, K., Willmott, P. & Dewhurst, M. (2017). A
Future that Works: Automation, Employment, and Productivity. McKinsey & Company.
Mazzucato, M. (2018) Mission-oriented innovation policies: challenges and opportunities Industrial
and Corporate Change, 27(5): 803815,
Mazzucato, M. & Perez, C. (2014). Innovation as Growth Policy: the challenge for Europe, SPRU
Working Paper Series, SWPS 2014D13, SPRU, University of Sussex.
McGuinness, S., Pouliakas, K. & Redmond, P. (2018). Skills mismatch: concepts, measurement and
policy approaches. Journal of Economic Surveys, 32(4): 9851015.
McLoughlin, I. & Clark, J. (1994). Technological Change at Work. Milton Keynes: Open University
Meil, P. & Kirov, V. (2017). Introduction: Policy Implications of Virtual Work. In P. Meil & N. Kirov
(eds.). (2017). Policy Implications of Virtual Work. London: Palgrave Macmillan.
Measuring Job Quality Working Group (2018) Measuring Good Work, Dunfirmline: Carnegie UK
Milana, M., Webb, S., Holford, J., Waller, R. & Jarvis, P. (eds.). (2018). Palgrave International
Handbook on Adult and Lifelong Education and Learning. London: Palgrave Macmillan.
Milkman, R. (1998). ‘The New American Workplace’ in P. Thompson & C. Warhurst (eds)
Workplaces of the Future, London: Macmillan.
Moore, J. F. (1993). Predators and Prey: A New Ecology of Competition. Harvard Business Review,
71: 30-76.
Mournier, A. (2001). The three logics of skill in French literature. Sydney: BVET.
Morel, N. Palier, B. & Palme, J. (eds. (2012). Towards a social investment state? Bristol: Policy Press.
Muñoz de Bustillo, R., Fernández-Macías, E., Esteve, F. & Antón, J. (2011). Measuring more than
money: the social economics of job quality. Cheltenham: Edward Elgar.
Muro, M., Maxim, R. & Whiton, J. (2019). Automation and Artificial Intelligence: How machines are
affecting people and place. Washington, DC: Brookings Institute.
Nelson, R.R. & Winter, S.G. (1982). An Evolutionary Theory of Economic Change. Cambridge,
Massachusetts: The Belknap Press.
Nelson, R.R. (1994). The Co-evolution of Technology, Industrial Structure, and Supporting
Institutions. Industrial and Corporate Change, 3(1): 4763.
Nickson, D. Warhurst, C., Commander, J., Hurrell, S. & Cullen, A-M. (2012). Soft Skills and
Employability: Evidence from UK Retail. Economic and Industrial Democracy, 33(1): 56-84.
Noble, D.F. (1984). Forces of production: A social history of industrial production. New York: Knopf.
OECD (2011) Towards and OECD Skills Strategy. Paris: OECD.
OECD. (2014). Innovation-driven Growth in Regions: The Role of Smart Specialisation. Paris: OECD
OECD. (2015). New skills for the digital economy - Measuring the demand and supply of ICT skills.
OECD Digital Economy Papers no. 258. Paris: OECD.
OECD. (2017). OECD Guidelines on Measuring the Quality of the Working Environment. Paris: OECD.
OECD. (2018a). Online work in OECD countries. Policy Brief on the Future of Work. Paris: OECD.
OECD. (2018b). Putting faces on the jobs at risk of automation. Policy Brief on the Future of Work.
Paris: OECD.
OECD. (2018c). Job creation and local economic development 2018: preparing for the future of work.
Paris: OECD.
OECD. (2018d). Good jobs for all in a chaging world of work. Paris: OECD.
Oeij, P. R. A., Rus, D. & Pot, F.D. (eds.) (2017). Workplace Innovation. Theory, Research and Practice.
Cham: Springer International Publishing.
Oeij, P.R.A. & Dhondt, S. (2017). Theoretical Approaches Supporting Workplace Innovation. In
P.R.A., Oeij, D. Rus & F. Pot (Eds.), Workplace Innovation - Theory, Research and Practice (pp. 63-
78). Cham: Springer Nature. Aligning Perspectives on Health, Safety and Well- Being.
Okay-Somerville, B. and D. Scholarios (2013), “Shades of grey: Understanding job quality in
emerging graduate occupations”, Human Relations, 66(4): 555-585,
Osterman, P. (2018). In search of the High Road: meaning and evidence’. ILR Review, 71(1), January,
Parmiggiani, A., Randazzo, M., Natale, L., Metta, G. (2014). An alternative approach to robot safety
IEEE International Conference On Intelligent Robots and Systems. 484-489. DOI:
Perez, C. (2002). Technological Revolutions and Financial Capital: The Dynamics of Bubbles and
Golden Ages. Cheltenham: Edward Elgar.
Perez, C. (2010). Technological revolutions and techno-economic paradigms. Cambridge Journal
of Economics, 34(1): 185202.
Pesole, A., Urzí Brancati, M.C, Fernández-Macías, E., Biagi, F. & González Vázquez, I. (2018).
Platform Workers in Europe: Evidence from the COLLEEM Survey, JRC Science for Policy report,
Pfeiffer, S. (2016) Robots, Industry4.0 and Humans, or Why Assembly Work Is More than Routine
Work Societies, 6(2), 16, 26 p.
Pintelon L., Muchiri P. (2009) Safety and Maintenance. In: Ben-Daya M., Duffuaa S., Raouf A.,
Knezevic J., Ait-Kadi D. (eds) Handbook of Maintenance Management and Engineering. London:
Pisano, P., Pironti, M. & Rieple, A. (2015). Identify innovative business models: can innovative
business models enable players to react to ongoing or unpredictable trends? Entrepreneurship
Research Journal, 5(3): 181-199.
Reardon, C., Tan, H. Kannan, B., DeRose, L. (2015) Towards Safe Robot-Human Collaboration
Systems using Human Pose Detection IEEE Conference on Technologies for Practical Robot
AM.pplications, TePRA, August, 7219658, 1-6 (DOI: 10.1109/TePRA.2015.7219658)
Rifkin, J. (2012). The Third Industrial Revolution. Basingstoke: Palgrave Macmillan.
Rodrik D, Sabel C. Building a Good Jobs Economy. Working Paper. Copy at
Rothschild, M., Stiglitz, J. (1976) Equilibrium in Competitive Insurance Markets: An Essay on the
Economics of Imperfect Information. Quarterly Journal of Economics, 90(4): 629-649.
Rubenson, K. & Desjardins, R. (2009). The impact of welfare state regimes on barriers to
participation in adult education. A bounded agency model. Adult Education Quarterly, 59(3): 187-
Russell, M.G., Still, K., Huhtamäki, J., Yu, C. & Rubens, N. (2011). Transforming Innovation
Ecosystems through Shared Vision and Network Orchestration. Paper 81.00. S1.1 History and
conditions for success. Conference paper, Triple Helix IX International Conference: “Silicon Valley:
Global Model or Unique Anomaly?, California: University of Stanford, 21 p.
Saez, E. & Zucman, G. (2019). The Triumph of Injustice: How the Rich Dodge Taxes and How to Make
Them Pay, New York (US): WW Norton & Co..
Sarkar, S. (2017). ‘Employment polarisation and over-education in Germany, Spain, Sweden and UK’,
Empirica, 44: 435-463.
Schor, J. & Fitzmaurice, C. (2015). Collaborating and Connecting: The Emergence of a Sharing
Economy. In L. Reisch & J. Thogersen. (eds). Handbook on Research on Sustainable Consumption.
Cheltenham: Edward Elgar.
Schumpeter, J.A. (1911). The Theory of Economic Development: an Inquiry into Profits, Capital,
Credit, Interest, and the Business Cycle. Cambridge, Massachusetts: Harvard University Press.
Schumpeter, J.A. (1939). Business cycles: a theoretical, historical, and statistical analysis of the
capitalist process. New York: McGraw Hill.
Schwab, K. (2016). The Fourth Industrial Revolution. London: Penguin.
Sitter, L. U. de, Hertog, J. F. & Dankbaar, B. (1997). From complex organizations with simple jobs to
simple organizations with complex jobs. Human Relations, 50(5): 497-536.
Skills Australia (2012). Better use of skills, better outcomes: A research report on skills utilisation in
Australia, Canberra: Skills Australia.
Soros, G. (2018)