BookPDF Available

European legal framework for “digital labour platforms”

Authors:

Abstract and Figures

This report maps a kaleidoscopic array of platform-mediated working arrangements, by clustering the findings into three main subsets (passenger transport services, professional crowdsourcing, on-demand work at the client’s premises). Many initiatives taken by the European institutions and aimed at promoting decent work in the collaborative economy are analysed including (i) the European Commission’s Communication 356/2016, (ii) the principles enshrined in the European Pillar of Social Rights, and (iii) the ruling by the European Court of Justice on the nature of the service provided by Uber. After exploring the existing legal framework in several European countries, this study goes into the issue of the legal status of platform–based or –mediated workers by analysing what is at stake in pending litigations on the proper classification. In the end, this report is meant to contrast the sense that new realities of work have outgrown legal concepts. The application of existing regulation must be reinforced, in order to avoid the risk that platform workers are considered by default as falling in a normative vacuum. In the end, creating a level playing field between the traditional and the digitally-enabled companies is the only way to reap full benefits of the on-going digital transformation.
Content may be subject to copyright.
European legal framework
for
“digital labour platforms”
De Stefano, V.
Al
oisi, A.
2018
This publication is a report by the Joint Research Centre (JRC), the European Commission’s science and
knowledge service. It aims to provide evidence-based scientific support to the European policymaking process.
The scientific output expressed does not imply a policy position of the European Commission. Neither the
European Commission nor any person acting on behalf of the Commission is responsible for the use that might
be made of this publication.
Contact information
Name: Valerio de Stefano
Email: valerio.destefano@kuleuven.be
JRC Science Hub
https://ec.europa.eu/jrc
JRC112243
PDF ISBN 978-92-79-94131-3
doi:10.2760/78590
Luxembourg: Publications Office of the European Union, 2018
© European Union, 2018
The reuse policy of the European Commission is implemented by Commission Decision 2011/833/EU of 12
December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Reuse is authorised,
provided the source of the document is acknowledged and its original meaning or message is not distorted. The
European Commission shall not be liable for any consequence stemming from the reuse. For any use or
reproduction of photos or other material that is not owned by the EU, permission must be sought directly from
the copyright holders.
All content © European Union 2018, except Cover image, by Rawpixel.com, © Adobe Stock, 2018
How to cite this report: De Stefano V. and Aloisi A., European Legal framework for digital labour platforms,
European Commission, Luxembourg, 2018, ISBN 978-92-79-94131-3, doi:10.2760/78590, JRC112243.
i
Contents
Foreword .............................................................................................................. 2
Acknowledgements ................................................................................................ 3
Abstract ............................................................................................................... 4
Executive summary ............................................................................................... 5
1 A taxonomy of digital labour platforms. ................................................................ 7
1.1 Crowdsourcing and work on-demand via platform. ......................................... 11
2 Understanding the platforms’ business models. ................................................... 14
2.1 An in-depth analysis on working conditions in three sub-sectors: a multi-stage
analysis. ......................................................................................................... 16
2.1.1 Access to the platform and registration. ................................................ 17
2.1.2 Selection process and hiring. ............................................................... 18
2.1.3 Command power and performance execution. ....................................... 19
2.1.4 Monitoring and rating (and deactivation). ............................................. 20
2.1.5 Payment rewards for completed tasks. ................................................. 22
3 Overview of the legal frameworks applying to labour platforms. ............................ 25
3.1 (European and) national requirements to comply with. ................................... 26
3.1.1 Legal frameworks for casual work: on-call work, voucher-based, and zero-
hours contracts. .......................................................................................... 32
3.1.2 Sectorial analysis. Passenger transport service. ..................................... 35
3.1.3 Sectorial analysis. Online crowdsourcing. .............................................. 38
3.1.3.1 Platforms as temporary work agencies, an “indirect” approach. ......... 39
3.1.4 Sectorial analysis. Work on demand via platform (at the household’s
premises). .................................................................................................. 40
4 Issues at stake and potential outcomes from key litigation cases. .......................... 41
4.1 Classification of workers: a gateway to protections. ....................................... 44
5 Final remarks and policy recommendations. ........................................................ 53
References ......................................................................................................... 55
List of boxes ....................................................................................................... 64
List of figures ...................................................................................................... 65
List of tables ....................................................................................................... 66
2
Foreword
The European labour market is experiencing radical changes, fuelled by digitalisation,
flexibilisation, restructuring, and demographic dynamics. Moreover, the digital economy
is changing societies and altering socio-economic interactions. Understandably, all these
trends have a strong impact on the organisation of work and, above all, on the
relationships between employers and employees (or between workers and clients). More
recently, a spotlight has been shone on the phenomenon of working in the so-called
“collaborative economy”, described as the most visible portion of a broader trend towards
de-standardisation of employment relationships and de-mutualisation of risk.
In a nutshell, labour platforms allow individuals, families or companies in need of a
service to hire a worker who is willing to offer the relevant activity, whether manual or
creative, under the guise of enhanced flexibility. By taking into account this significant
heterogeneity, which may entail intense juridical implications, a common business model
could be traced consisting in the “instant” matching of demand and supply of labour,
facilitated by digital tools (mostly apps on smartphones and online platforms) that make
it easy to manage a large and “low-cost” workforce, lowering transaction costs, barriers
to entry and information asymmetries.
While it is correct to maintain that, at least in theory, digital labour platforms may
expand competitiveness, increase choice and create growth opportunities, a series of
concerns has been raised on the potential erosion of rights, and the non-compliance with
employment law standards. Moreover, digitally mediated working templates blur
boundaries between traditional classifications such as work and rest periods, amateurism
and professionalism, subordination and autonomy. This model touches on the foundation
of traditional social protection and social security systems.
From a regulatory perspective, platform labour represents a “moving target”. Building on
the results of the wide-ranging investigation into concrete working conditions in three
selected subsectors, namely passenger transport services, professional crowdsourcing,
on-demand work at the client’s premises, the crucial question that needs answering is
whether the existing legal frameworks provide adequate responses to the current
(digital) challenges. To gain a better understanding, standard methods of social science
research and legal analysis will be combined to outline regulatory and contractual
templates, which can easily accommodate modern organisational patterns.
Accordingly, the aim of this study is to contribute to the comprehensive research on
digital labour platforms by investigating the existing legal frameworks and understanding
related challenges for policy makers. Ambitious expectations are being pinned new
technologies making work and lives considerably innovative and cannot be let down. Still
employment rights must not be sacrificed on the altar of flexibility. The goal, indeed, is to
develop sustainable digital ecosystems and social institutions where high-quality
employment and competitiveness mutually reinforce.
3
Acknowledgements
Our appreciation to Professor Stefano Liebman, Dean of the School of Law at Bocconi
University, Milano, Professor Miriam A. Cherry, Co-Director of the William C. Wefel Center
for Employment Law at Saint Louis University School of Law, and Janine Berg of the
International Labour Office, who have been extremely helpful with this report. Many
thanks also to Mathias Wouters, Institute for Labour Law, KU Leuven. The usual
disclaimer applies.
Authors
Valerio De Stefano, Antonio Aloisi.
4
Abstract
This report maps a kaleidoscopic array of platform-mediated working arrangements, by
clustering the findings into three main subsets (passenger transport services,
professional crowdsourcing, on-demand work at the client’s premises). Many initiatives
taken by the European institutions and aimed at promoting decent work in the
collaborative economy are analysed including (i) the European Commission’s
Communication 356/2016, (ii) the principles enshrined in the European Pillar of Social
Rights, and (iii) the ruling by the European Court of Justice on the nature of the service
provided by Uber. After exploring the existing legal framework in several European
countries, this study goes into the issue of the legal status of platform–based or
mediated workers by analysing what is at stake in pending litigations on the proper
classification. In the end, this report is meant to contrast the sense that new realities of
work have outgrown legal concepts. The application of existing regulation must be
reinforced, in order to avoid the risk that platform workers are considered by default as
falling in a normative vacuum. In the end, creating a level playing field between the
traditional and the digitally-enabled companies is the only way to reap full benefits of the
on-going digital transformation.
5
Executive summary
This research focuses on emerging digital platforms which enable the matching of labour
demand and supply, through highly efficient infrastructure connecting a vast number of
workers, employers and clients instantaneously. Such far-reaching vicissitudes hold great
perils, but also huge potential. The report intends to “normalise” the discourses
surrounding the digital transformation of work and the rise of non-standard forms of
employment, by contrasting the sense that new realities of work have outgrown legal
concepts.
Policy context
Only few European States have adopted specific regulations to address the numerous
issues stemming from the advent of the platform economy: the model is mercurial in
nature and a hefty intervention may provoke its premature asphyxiation. Accordingly,
before proposing one-size-fits-all or horizontal regulatory schemes, it would be useful to
validate the appropriateness of existing labour law categories, by delving into labour law
fitness to these new realities without ignoring the sheer heterogeneity of this
phenomenon. While specific legislative and regulatory responses to these issues lag
behind, the research reviews legislation as well as national practices with regard to
platform-facilitated arrangements and new forms of (web–based or –mediated) work.
Against this background, European initiatives aimed at promoting decent work in the
collaborative economy receiving a high level of research attention include: (i) the
European Commission’s Communication 356/2016, (ii) the principles enshrined in the
European Pillar of Social Rights, and (iii) the ruling by the European Court of Justice on
the nature of the service provided by the ride-hailing company Uber. Moreover, existing
regulatory schemes, ranging from European Directives on atypical employment to casual
work templates such as zero-hours or voucher-based contracts, are investigated.
Key conclusions
Labour platforms are quickly and partly redesigning the way work and firms are
organised, by developing at the fringes of regulation. Contrary to what is often said, it is
argued that the employment relationship and, more in general, current legal formats are
not undergoing an irreparable crisis. In brief, it is demonstrated how it is feasible to
temper the impulse to digital distinctiveness with actions to safeguard workers’ rights.
In addition to this, it cannot be underestimated that the lack of compliance with labour-
related, fiscal and social security duties constitutes platforms’ main competitive
advantage vis-à-vis their competitors. This regulatory arbitrage results in an aggravation
of existing conceptual tension and, what is worse, in an exacerbation of social
precariousness as platform workers have very limited access to labour protection.
Before proposing one-size-fits-all or horizontal regulatory schemes, it would be useful to
validate the appropriateness of existing labour law categories, by delving into labour law
fitness to these new realities without ignoring the sheer heterogeneity of this
phenomenon. Indeed, several tools could already be used to regulate platform–based or
–mediated working patterns.
Main findings
The (new) world of work is characterised by an increased tendency towards relationships
that are not based on direct employment contracts. Although the description of specific
circumstances does not allow for generalisations, a multi-stage scrutiny might provide a
complete picture of the question. To fully appreciate the variety of services that fall under
the notion of platform-mediated labour, this report maps a kaleidoscopic array of app-
distributed employment arrangements, by clustering the findings into three main subsets
(passenger transport services, professional crowdsourcing, on-demand work at the
client’s premises).
6
Scrutinising the concrete operation of the on-demand platforms could help scholars,
practitioners and policymakers in addressing more complex issues concerning the scope
of employment law and, broadly, the state of play of the social compact. The operation of
a set of European as well as global platforms will be described throughout five different
dimensions. In doing so, the relation between a worker and the platform is considered
along five main key phases: (i) access to the platform and registration, (ii) selection
process and hiring, (iii) performance execution and command power, (iv) rating and
ranking, monitoring power (and deactivation), (v) payment rewards for completed tasks.
This report goes into the issue of the legal status of platform-coordinated workers by
analysing what is at stake in pending litigation on the proper classification. Recent and
potential outcomes are sketched. Although firms engage with “providers”, arguably and
invariably classified as self-employed, on a hyper-volatile basis, many of them exert
some degree of managerial powers, albeit in a more sophisticated form, over the
transacted activity, while avoiding the obligations of direct employment.
Available research reveals pervasive directives, reinforced surveillance, constant
assessment, arbitrary disciplinary action and very little or no margin in deciding how to
complete a task. The model does not contain any entitlement such as overtime, paid
holiday leave, maternity leave, sickness payments and statutory minimum wages.
Furthermore, workers are excluded from fundamental principles and rights at work such
as freedom of association, collective bargaining or protection against discrimination or
unfair dismissal.
Therefore, this apparently new pattern has to be on the radar for several reasons. Firstly,
the “platformisation” of labour relations may reshape and revive traditional outsourcing
practices. Secondly, it could contribute to a formalisation of informal economy. Thirdly, it
might speed up the erosion process of both traditional categories and protections –
allowing companies to gain a competitive advantage over traditional competitors that
comply with labour law provisions. Thus, such hybrid arrangements call into question the
suitability and effectiveness of the current employment legislation.
The modalities in which digitally enabled work is performed differ much from one
another. While focusing on the “distribution channel” (i.e. platforms) is convenient for
academic purposes (even though it is not a legal criterion for distinction), regulating app-
mediated work as an entirely unique category deserving new notions, new concepts, new
regulations should be avoided: the design of specific legislation only targeting platforms
makes little sense, at least in the area of labour law and social security.
Therefore, reaffirming the major binary divide between workers genuinely self-employed
and those in an employment relationship would be helpful in the attempt to reduce
labour market segmentation and rising inequalities. It can be argued that platform-
mediated arrangements are particularly keen sites for investigating how classical legal
notions and formats can adapt to new dynamics. Understandably, another answer might
be closing legal loopholes that incentivise exploitative conducts by a marginal group of
deceitful companies.
Some old issues gain new attention in the age of digital transformation of work.
“Surgical” regulatory interventions shall help the collaborative economy companies to
improve their business model, building a new phase of “shared social responsibility”. In
this sense, the current European attitude is perceived as a fair balance between
supporting entrepreneurs’ confidence and implementing workers’ protections, but
considerable efforts need to be done in order to ensure a stable and sustainable future.
7
1 A taxonomy of digital labour platforms.
The European labour market has experienced radical changes in the last years, fuelled by
digitalisation, flexibilisation, restructuring, and demographic dynamics. Understandably,
all these trends have a strong impact on the organisation of work and, above all, on the
relationships between employers and employees (or between workers and clients). Whilst
creating potentially new opportunities for workers, such as a better-quality work-life
balance, casual arrangements might determine shortcomings, such as unstable or rapidly
changing schedules. Moreover, digitally mediated working templates blur boundaries
between traditional classifications such as work and rest periods, amateurism and
professionalism, subordination and autonomy. More recently, a spotlight has been shone
on the phenomenon of working in the so-called “collaborative economy”, described as the
most visible portion of a broader trend towards de-standardisation of employment
relationships and de-mutualisation of risk (
1
). While it is correct to maintain that, at least
in theory, digital labour platforms may expand competitiveness (but also competition),
increase choice and create growth opportunities, a series of concerns has been raised on
the potential erosion of rights, protections and social security. Though, researches have
mostly provided anecdotal evidence and knowledge is based on interviews and personal
first-hand experiences.
To begin with, navigating definitions is an intricate assignment when it comes to describe
digitally enabled forms of work in the so-called “platform economy”, known in everyday
parlance also as the “gig-economy” (
2
). At first glance, the list of labels grows day by
day, together with the number of businesses and people involved in this sector. In this
report, notwithstanding the nuances of meanings, “platform economy” (
3
), “on-demand
economy”, and “collaborative economy” will be used interchangeably (
4
). Whilst
representing seducing buzzwords, these formulas have reached a large consensus and
they are therefore well suited to indicate a rather specific economic segment. At the
same time, it has to be acknowledged that these labels cover various kinds of economic
actors that might have widely different normative and social implications, while sharing a
significant number of features. Therefore, referring to them as a monolithic category
causes many commentators and practitioners to be confused about what exactly is being
described or studied.
Due to difficulties in measuring the relevant ecosystem, there is still little statistics and
scant knowledge about its dimensions (
5
): different methodologies may result in
divergent assessment (
6
). Nevertheless, many analyses seem to point out unanimously
that the collaborative economy is constantly on the rise. This economic segment seems
to be growing by 25% a year. According to more generous estimates, its value in Europe
exceeds €20 billion (
7
). A very detailed document (
8
) finds that the platform economy
(
1
) Defined as the trend of “passing some of the risks of the enterprise onto the worker”. See F
REDMAN
S.
(2006), Precarious Norms for Precarious Workers in F
UDGE
J. and O
WENS
R. (Eds.), Precarious Work,
Women and the New Economy, Oxford, p. 187.
(
2
) The term “gig” evokes artists jumping from a concert to another.
(
3
) See Communication from the Commission to the European Parliament, the Council, the European Economic
and Social Committee and the Committee of the Regions, A European Agenda for the Collaborative
Economy, COM/2016/0356 final (hereinafter “the Communication”).
(
4
) Although the plethora of online platforms sees itself as a part of this trend, the label of “sharing economy” is
now disgraced for misrepresenting the reality, after a preliminary moment of excitement. For a complete
overview of definitions, see S
ELLONI
D. (2017), New Forms of Economies: Sharing Economy, Collaborative
Consumption, Peer-to-Peer Economy in E
AD
., CoDesign for Public-Interest Services, Cham, p. 15.
(
5
) Suffice here to highlight here the distinction between active and non-active accounts or the use of multiple
identities to register on different platforms.
(
6
) Consistent estimates are hard to come by. The value of the collaborative economy in the EU varies from
survey to survey, from report to report. For a review of many intrinsic problems, see H
ARDIE
M. (2016), The
feasibility of measuring the sharing economy, UK Office for National Statistics (describing “a number of
challenges faced when attempting to measure [this phenomenon], including classifying activit[ies],
capturing sharing activity between individuals and measuring non-monetary transactions”).
(
7
) Reports have revealed how online platforms are not generating sufficient income. Moreover, it has to be said
that “[v]ery few of the digital platform companies currently make profits, but the very high market
capitalisation of these companies obviously means that the stock market believes that they will do so
sometime in the future”. See E
UROFOUND
(2017), Non-standard forms of employment: Recent trends and
8
amounts to nearly €4 billion in revenues and has intermediated €28 billion of transactions
(85% of this value being gained by the providers/workers). According to a recent
Eurobarometer survey, 52% of all EU citizens are aware of the services offered by the
sharing economy (
9
). In addition to that, it should be said that participation in the
collaborative economy is “relatively small – but growing”: between 5% and 9% of
European citizens have already participated in this framework, making no more than
€1000 (the median of earning stands at around €300) (
10
). According to a report
commissioned by the European Commission (
11
), there would be approximately 100,000
active workers in the European platform economy, representing 0.05% of the total
workforce. Before venturing on the task of scrutinising platforms more in depth, it should
be borne in mind that their potential has not been all unleashed yet. More importantly,
workers do not address platforms only for amateur purposes; on the contrary, a
significant proportion relies on these intermediaries for their principal source of income,
according to a survey carried out by ILO on crowd-workers (
12
).
In considering the background, scholars have identified three main reasons explaining
the rise of the platform- or app-based workforce: (i) the need to cope with short-run
fluctuations on the demand side, (ii) the desire to reduce labour costs, (iii) the urgency
to meet market pressures on short-term results and efficiency (
13
). In order to grapple
with the constant evolution, this contribution will look into the atypical forms of work
which are channelled via digital infrastructures (
14
). Classifying models in this field has
attracted a lot of consideration, as recent studies on this topic have classified several
typologies of platforms as well as of digitally enabled work arrangements. Legal scholars
tend to “focus separately […] on online and offline workers, because their places of work
(remote versus face-to-face) and relationships with clients (telemediated versus direct)
create very different patterns of work, exposing them to different risks” (
15
). Setting a
detailed categorisation could help in selecting the most salient realities to be assessed
from a legal standpoint.
future prospects, Background paper for Estonian Presidency Conference ‘Future of Work: Making It e-Easy‘,
13-14 September 2017, p. 22.
(
8
) V
AUGHAN
R.
and
D
AVEIO
R. (2016), Assessing the size and presence of the collaborative economy in Europe,
PwC UK, Impulse paper for the European Commission. The report identifies five key sectors (peer-to-peer
accommodation, peer-to-peer transportation, on-demand households service, on-demand professional
services, collaborative finance). The study indicates that there are 275 collaborative economy platforms in
9 states (Belgium, France, Germany, Italy, Poland, Spain, Sweden, The Netherlands, and UK).
(
9
) E
UROPEAN
C
OMMISSION
(2016), The use of collaborative platforms, Flash Eurobarometer, No. 438. Survey
requested by the European Commission, Directorate-General for Internal Market, Industry,
Entrepreneurship and SMEs and co-ordinated by the Directorate-General for Communication.
(
10
) A preliminary survey implemented in the UK, Sweden, Germany, Austria and the Netherlands in the first
two quarters of 2016 will be expanded in November 2017.
(
11
) D
E
G
ROEN
W.
P. and M
ASELLI
I.
(2015), The Impact of the Collaborative Economy on the Labour Market,
CEPS Special Report No. 138. The empirical strategy based on the frequency of Google searches for words
related to online platforms is the same carried out in
H
ARRIS
S.
D. and K
RUEGER
A.
B. (2015), A Proposal for
Modernizing Labor Laws for Twenty-First-Century Work: The “Independent Worker”, The Hamilton Project.
It is worth noting, however, that the estimate is “significantly less than the 0.4% to 1% of employees that
is assumed to be participating in the US”. Simply put, the platforms appeared later in the EU and “[took]
more time to develop due to fragmented markets and regulation, as well as the fact that labour is more
protected and sector concerns may be more regulated”, see D
E
G
ROEN
W.
P., L
ENAERTS
K., B
OSC
R. and
P
AQUIER
F. (2017), Impact of Digitalisation and the On-Demand Economy on Labour Markets and the
Consequences for Employment and Industrial Relations, Study prepared for the European Economic and
Social Committee, Brussels, p. 9, n. 4.
(
12
) See B
ERG
J. (2016), Income Security in the On-Demand Economy: Findings and Policy Lessons from a
Survey of Crowdworkers in Comp. Lab. L. & Pol’y J., 37(3) (reporting the main results of a survey which
shows that almost 40% of respondents – out of a sample of crowdworkers on the Amazon Mechanical Turk
and on Crowdflower – rely on platforms as their principal source of sustenance).
(
13
) D
OKKA
J.,
M
UNFORF
M.
&
S
CHANZENBACH
D.
W.
(2015), Workers and the Online Gig Economy, The Hamilton
Project.
(
14
) G
ILLESPIE
A.,
R
ICHARDSON
R.,
V
ALENDUC
G. et al. (1999), Technology Induced Atypical Work-Forms. Report for
the Office of technology assessment of the European Parliament (STOA), Brussels.
(
15
) H
UWS
U.,
S
PENCER
N.
H. and J
OYCE
S. (2016), Crowd Work in Europe. Preliminary Results from a Survey in
the UK, Sweden, Germany, Austria and The Netherlands, Foundation for European Progressive Studies.
Nonetheless, outlining an inflexible dichotomy between these two forms of work is not convincing. See D
E
S
TEFANO
V. (2016), Introduction: Crowdsourcing, the Gig-Economy, and the Law in Comp. Lab. L. & Pol’y J.,
37(3), p. 461.
9
Admittedly, many commentators agree on dividing this broad array into two main
models: (a) crowd-sourcing, and (b) work on-demand via platform or app (
16
). In an
attempt to resolve this definitional dilemma, a preliminary distinction between
crowdsourcing and work on-demand can be drawn by taking into account the place of
performance of work and considering the mechanisms through which work is requested
or obtained. At the same time, different approaches in emphasizing the major
characteristics of these schemes may lead to alternative categorisation. Without going
too far on this point, platforms may be classified taking into account several
characteristics: (i) the dimension of the platforms, distinguishing between a global or
local presence, (ii) the site of the execution, differentiating between online tasks and
“real world” tasks, (iii) the content of the “gigs”, distinguishing between creative, routine
or manual jobs, (v) the service offered, distinguishing between “task specific” and
“generalist” platforms, (vi) the nature of the performances, distinguishing between “low-
skill” and “high-skill” activities (
17
), (vii) the way of adjudication (contest vs.
procurement), (viii) the system of payment (free bid versus fixed rate). Besides, looking
at the location, work on-demand can be split into transport services and household
services.
In particular, this report will concentrate its attention on three main subgroups of
concrete working performances in the platform economy: (i) passenger transport
services, (ii) professional tasks completed online (
18
), (iii) manual services carried out on
the household’s premises, whether domestic or commercial (See Figure 1 for a complete
picture). The research will review national practices and legal cases as well as legislations
introduced with regard to labour platforms and new forms of (web-based or -
intermediated) work. In doing so, attention will be drawn to eight European states,
namely Belgium, France, Germany, Italy, Ireland, Spain, Sweden and United Kingdom,
selected for language reasons and availability of data as well as knowledge about
legislative, judicial and entrepreneurial or collective initiatives. In the name of
completeness, the very first distinction is between labour platforms and platforms that
facilitate access to goods, property and capital (
19
). Among the latters, one can list
capital platforms (such as the British Funding Circle (
20
)), pure rental services (such as
AirBnB, HomeAway (
21
) or Turo (
22
)), peer-to-peer car sharing services (such as the
French BlaBlaCar (
23
)) and peer-to-peer marketplace for exchanging goods (such as the
Spanish Grownies (
24
)). Understandably, this analysis will not address this type of
platforms.
(
16
) B
ERG
J. and D
E
S
TEFANO
V. (2015, July 10), Regulating work in the ‘gig economy’, retrieved from
https://goo.gl/cs3tXp.
(
17
) Z
ITTRAIN
J. (2009, July 12), The Internet creates a new kind of sweatshop, retrieved from
https://goo.gl/UBUfjs (describing “a pyramid of sorts, with services designed to tap serious (and rare)
smarts at the top, and others to enlist anyone with a brain wave at the bottom”).
(
18
) See H
UWS
U. (2009) The making of a cybertariat? Virtual work in a real world in Socialist Register, 37, p. 2
(claiming that the formula non-manual work” would have denied “the physical reality of pounding a
keyboard all day”).
(
19
) D
RAHOKOUPIL
J.
and
F
ABO
B.
(2016), The platform economy and the disruption of the employment
relationship, ETUI Policy Brief, No. 5., p. 2. See also F
ARRELL
D. and G
REIG
F. (2016), Paychecks, Paydays,
and the Online Platform Economy: Big Data on Income Volatility, JP Morgan Chase Institute. Also in this
pattern, “there is some work to be done, such as driving the vehicle by the car owner or the guest’s
accommodation by flat owners”. A different dividing line could be mere sharing of costs vs. profit-seeking
motive. See also B
OCK
A.-K.,
B
ONTOUX
L.,
F
IGUEIREDO DO
N
ASCIMENTO
S.,
S
ZCZEPANIKOVA
A.
(2016), The future
of the EU collaborative economy Using scenarios to explore future implications for employment, JRC
Science for Policy Report.
(
20
) Funding Circle is a platform connecting small and medium enterprises with lenders and investors.
(
21
) AirBnB is probably the most known hospitality service. It matches hosts and guests enabling the lease or
rent of short-term lodging. It was founded in the U.S.A. HomeAway is a vacation rental platform.
(
22
) Turo, formerly RelayRides, is a company that operates a peer-to-peer car-sharing marketplace. It allows
private car owners to rent out their vehicles via an online platform.
(
23
) BlaBlaCar is a platform connecting drivers with idle seats with travellers or passengers paying a
contribution to the costs for long-distance or city-to-city routes. Founded in France. See A
URIEMMA
S.
(2017), Impresa, lavoro e subordinazione digitale al vaglio della giurisprudenza in RGL, I, p. 281
(explaining how its business model can be distinguished from other services in the same sector by virtue of
the cost-sharing mechanism).
(
24
) Grownies is a peer-to-peer marketplace for goods and clothes for young children. Founded in Spain.
10
Figure 1. Typologies of (labour) platforms.
Source: Authors’ own elaboration.
In order to get a better understanding of the “new” employment trends emerging across
Europe, a preliminary remark is needed. It is nearly impossible to enclose a very complex
phenomenon into rigid schemes, at least in the embryonic stages of platform
development. Each platform is organised in such a peculiar way that studying, and even
regulating, this issue with a “one-size-fits-all” approach results unfeasible. By taking into
account this significant heterogeneity, which may entail intense juridical implications, a
common business model could be traced consisting in the “instant” matching of demand
and supply of labour, facilitated by digital systems (mostly apps on smartphones and
online platforms) that make it easy to manage a large and “low-cost” workforce. In a
nutshell, platforms allow individuals, families or companies in need of a service to hire a
worker who is willing to offer the relevant activity, under the guise of enhanced flexibility
(“Be Your Own Boss” was one of Uber’s mottos). Dissimilarities among companies, as will
be demonstrated in the following paragraphs, go much beyond.
The aim of this study is to contribute to the comprehensive research on digital labour
platforms understanding legal frameworks and related challenges for policy makers. In
particular, after defining the main categories of platforms, this report will provide an
overview of regulatory frames applying to them. The operation of a set of European as
well as global platforms will be described throughout five different phases. Furthermore,
labour law issues arisen from these “atypical” or “casual” work arrangements will be
analysed. To this end, the attention will be devoted to potential or on-going legal
controversies regarding classification of workers which in turn impacts on working
conditions, fair treatment, and collective rights in these legal regimes (
25
). After looking
at those requirements that platform companies have to comply with, measures taken by
different Member States will be discussed. This paper concludes by illustrating potential
outcomes from litigation cases impacting on the employment status issue and, briefly,
policy implications at stake.
(
25
) See also Commission Staff Working Document, Communication from the Commission to the European
Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions –
a European Agenda for the Collaborative Economy.
11
1.1 Crowdsourcing and work on-demand via platform.
Clearing up any misunderstanding on this issue, it is essential to acknowledge the
professional nature of the performances rendered through labour platforms (
26
). There
has been a sustained attempt to present “gig work” as something related to generosity,
altruism or rather volunteerism, thus excluded from the field of labour law. Irrespective
of the reason behind the decision to engage with these economic activities, the
accomplishment of contingent tasks shall not be considered merely as an act of good
neighbourly relations or a spare-time activity. Clearly, this “understatement” cannot
apply to duties such as driving car, cleaning flats, translating documents, drawing up a
budget and completing secretarial work (
27
), just because they are “intermediated”
through the Internet. To this extent, a clean-up exercise is due. Using euphemisms such
as “Runners”, “Riders”, “Taskers”, “Clickers”, “Authors”, “Hourlies” or even “Friends” is a
clear attempt to hide the fact that human labour is at stake. The aim of this “newspeak”
is to counterfeit the reality or, at least, to gather arguments to be sustained in a
potential lawsuit.
After an introductory comment, it has to be pointed out that professional crowdwork (or
“crowdsourcing” (
28
)) includes services completed solely on computer and delivered
remotely, encompassing a wide variety of activities, clerical or repetitive (so-called “click
work”) as well as creative and intellectual ones, such as conceiving marketing campaigns
or proofreading academic contributions (
29
). The list of dominant performances, also
labelled as “human intelligence tasks” (HITs on Amazon Mechanical Turk), is rather long,
ranging “from data entry and admin work over graphic design and coding to legal and
business consulting” (
30
). The pattern involves routine performances (ClickWorker is an
example) executed from the provider’s house or from any place equipped with Wi-Fi
and delivered electronically to a “crowdsourcer” (
31
). These activities can be simply
performed by humans, but are still too complicated to be accomplished by computer
algorithms. Platforms can decomponetise work into a broad range of tiny jobs that can be
disseminated, claimed and performed (
32
).
Online tasks call for special mention. At the moment, consumers as well as policy makers
end up in overlooking the fact that a large portion of online activities is run by humans.
To this extent, Clickworker’s presentation is disarming: “As much as modern computing
power has increased, there are still many things only humans can do. (…) Clickworker
gets tasks done that computers can’t process, won’t process, (because the cost of
programming or equipment is too high) or those you can’t do because you don’t have
enough human resources to complete the project on time and on budget” (
33
). As it may
be clear, humans are behind the digital curtain. Artificial intelligence has been described,
(
26
) A detailed taxonomy of the online labour market is developed in C
ODAGNONE
C.,
A
BADIE
F.
and
B
IAGI
F.
(2016), The Passions and the Interests: Unpacking the ‘Sharing Economy’, JRC Science Policy Report,
Luxembourg.
(
27
) L
IEBMAN
W.
B. and L
YUBARSKY
A. (2016), Crowdwork, the Law, and the Future of Work, in Perspectives on
Work.
(
28
) The journalist Jeff Howe coined the term “crowdsourcing” in 2006. See H
OWE
J. (2006, June), The Rise of
Crowdsourcing, retrieved from https://www.wired.com/2006/06/crowds/ (describing “the act of taking a
job traditionally performed by a designated agent (usually an employee) and outsourcing it to an
undefined, generally large group of people in the form of an open call”).
(
29
) The most common activities on Amazon Mechanical Turk include, but are not limited to: building and
cleaning databases, testing software, validating search results, labelling items, transcribing audio clips and
recordings, harvesting email, editing documents, conducting surveys, recognising irony, placing ad on
videos, and detecting obscene images.
(
30
) OECD
(2016), New forms of work in the digital economy, Digital Economy Papers, No. 260, available at:
https://goo.gl/0b3c4t, p. 17.
(
31
) See E
UROFOUND
and I
NTERNATIONAL
L
ABOUR
O
FFICE
(2017), Working anytime, anywhere: The effects on the
world of work, Luxembourg – Geneva.
(
32
) M
ILLAND
K. (2017), Slave to the keyboard: the broken promises of the gig economy in Transfer, 23(2), p.
229. See also C
EFKIN
M.,
A
NYA
O.
and
M
OORE
R. (2014), A perfect storm? Reimagining work in the era of the
end of the job in Ethnographic Praxis in Industry Conference Proceedings, 1, p. 4.
(
33
) Clickworker’s presentation.
12
in a way, as anything but “artificial”
(
34
), since it is powered by human labour which can
be employed beyond the capabilities of current technology. According to the World Bank,
“microwork” is becoming a significant part of digital work, allowing “broader access to
specialized skills, more flexible and faster hiring processes, and 24-hour productivity”
(
35
). It consists in executing a variety of tasks including recommendations for e-
commerce website, detecting inappropriate materials online (from hate speech to
pornography), providing customer assistance via chats and forums. Moderation of user-
generated content on websites is another widespread application. In this case, legal
implications could be even more serious (i.e. child labour hidden behind avatars, the risk
of forced labour and “gold farming” in developing countries’ sweatshops (
36
)), in addition
to concerns regarding health and safety measures. Hence, crowdsourcing is supposed to
have two different upshots: on one hand, it provides the possibility to perform jobs that
were not previously accessible locally or to specific groups of workers. On the other, it
leads to a global mobility of existing jobs. Whether these dynamics may have an impact
on job creation or rather destruction is still not evident, further research is needed on
this specific point.
Coming now to the second model, the expression “on-demand work via platform or app
refers to services executed locally (mostly on the customer’s premises), such as delivery
(Foodora), maintenance (Etece.es), handyman (TaskRunner), cleaning services
(Helpling), and baby-sitting, to mention only the most important (
37
). Transportation
services, although not precisely delivered at the customer’s premises, belong to this
category
38
. Anyway, literature has focused mostly on the ride-hailing sector, as this one
represents the most blatant manifestation of risk and opportunities, and exemplifies the
dominant double narrative concerning platform-based economy. Needless to say, Uber
a very metonym for the sector (
39
) has inveigled its way into popular culture with
shocking speed, giving the birth even to a new verb (
40
).
It would be fair to say that most of these workers are excluded by some, or even the
entirety, of the essential protections due to employees (subordinate workers), such as
sick or holiday leave, full insurance, pension, superannuation or similar scheme, or
minimum rates of pay and working time. In addition to this, costs associated with
equipment, maintenance, reparations are at the provider’s own charge: no
reimbursements for these expenses are due. More often than not, the “participation
(
34
) P
RASSL
J. (forthcoming), Humans as a service, Oxford. In an open acknowledgment in 2013, former Waymo
and Uber engineer Anthony Levandowski, with such complete disregard for human dignity, presented a
Google team in India made up of what he called “human robots”, who were cataloguing images from Street
View application. See B
RADSHAW
T. (2017, July 9), Self-driving cars prove to be labour-intensive for
humans, retrieved from https://on.ft.com/2uHBiJq.
(
35
) See K
UEK
S.
C.,
P
ARADI
-G
UILFORD
C.,
F
AYOMI
T.,
I
MAIZUMI
S.,
I
PEIROTIS
P.,
P
INA
P.
and
S
INGH
M. (2015), The
global opportunity in online outsourcing, Washington, DC, p. 3 (explaining how “[i]ndustry experts suggest
that these firms currently have combined annual global gross services revenue of about $120 million;
together they form about 80 per cent of the microwork market”).
(
36
) C
HERRY
M. (2016), Virtual work and invisible labor in C
RAIN
M.,
P
OSTER
W. and C
HERRY
M. (Eds.) (2016),
Invisible Labor: Hidden Work in the Contemporary World, Oakland, CA, p. 83 (arguing that “ensuring non-
coerced or non-child sources of labor are just as salient here as they are in other forms of labor”). See also
C
ASILLI
A. (2016), Is There a Global Digital Labor Culture? Marginalization of Work, Global Inequalities, and
Coloniality, Paper presented at the 2
nd
symposium of the Project for Advanced Research in Global
Communication (PARGC), Philadelphia.
(
37
) The distinction is not clear-cut, as other platforms, such as Freelancer, allow selecting workers for both
virtual and real world services. From coding and design to “help with my garden maintenance” and “clean
my house or business”. See https://www.freelancer.com/freelancers/.
(
38
)
A
DAM
D.,
B
REMERMANN
M.,
D
URAN
J.,
F
ONTANAROSA
F.,
K
RAEMER
B.,
W
ESTPHAL
H.,
K
UNERT
A.
and
T
ÖNNES
L
ÖNNROOS
L.
(2016),
Digitalisation and working life: lessons from the Uber cases around Europe, EurWORK, European
Observatory of Working Life.
(
39
) Uber’s impressive success further inspires other platform-based companies. It very often happens to come
across passionate entrepreneurs launching the Uber but for X”. It is indeed useful as an expository
example, due to the “replicability” of its business model. Nevertheless, caution has to be exercised towards
the ride-hailing company
(
40
) In France the word Ubérisation has become a synonym of casualisation. It was said, and quite rightly too,
that “[f]ew companies offer something so popular that their name becomes a verb”, see Uberworld (2016,
September 3), retrieved from https://econ.st/2bKeQVA. Indeed, the company has become a dominant
metonym for the decline of labour-intensive industries, and hence for “reinventing” jobs.
13
agreement” (i.e. non-negotiable contract) specifies that the worker performs his duty as
an independent contractor (self-employed worker). A very popular clause points out that
providers use the platform at their own risk. Undervaluing the fact that most of labour
related, fiscal and social security costs are not paid is detrimental, all the more so
because digital platforms sometimes exert strong prerogatives involving persistent
monitoring power, unilateral arrangement of terms and condition and deactivation
privilege, as explained in the following section.
Therefore, this apparently new pattern has to be on the radar for several reasons. Firstly,
the “platformisation” of labour relations may reshape and revive traditional outsourcing
practices. Secondly, it should contribute to a formalisation of informal economy, by
“replacing word-of-mouth methods of finding work, and unrecorded cash-in-hand
payments by traceable online payments” (
41
). Thirdly, it might speed up the erosion
process of both traditional categories and protections allowing companies to gain a
regulatory arbitrage over traditional competitors that comply with labour law provisions.
(
41
) H
UWS
U. (2016, June 1), Logged In, retrieved from https://goo.gl/ik6BWL.
14
2 Understanding the platforms’ business models.
Having set the perimeter of this phenomenon, the platform’s business model will be now
scrutinised thoroughly. In this regard, in 2016 the European Community took the stance
defining the scope of the notion of collaborative economy as “business models where
activities are facilitated by collaborative platforms that create an open marketplace for
temporary usage of goods or services often provided by private individuals” (
42
). In most
cases, this model consists in chopping up a multipart work into its smallest components
and submitting each of them to always available and geographically dispersed “legions”
of workers. Many platforms for clerical services are operating globally as they can profit
from the almost universal knowledge of the English language; others instead are
mediating casual work locally, sometimes contributing to reduce undeclared employment
(incredibly widespread in certain sectors such as janitorial services). Understanding the
platforms’ business model might help ascertaining whether workers are employed or self-
employed, as explained in section 4.
As already flagged above, looking at labour platforms as a unified whole could be
inappropriate as well as misleading. Despite the common features, such as the enabling
role played by ICT tools, much dissimilarity can be traced. In this paragraph, an attempt
will be made to highlight shared hallmarks of this autonomous but heterogeneous
archetype. As far as the business model is concerned, the “digital matching firms” at
hand can be characterised focusing on a list of principal factors. First and foremost, the
massive use of advanced information technology, typically the combination of widespread
broadband, user-friendly digital application and increasingly effective features to facilitate
transactions and keep the organisation lean; then the reliance on customer-based
feedback systems for quality check as well as the reliance on workers using their own
equipment (be they personal computer, bicycles or cars, whether leased or owned) to
provide a service (
43
). As mentioned, the interrelationships between actors could be
described as triangular, being the platform (which controls intellectual property rights
and governance) a connecting system between buyers (“requesters”, according to the
internal jargon) and workers (“sellers” or “providers”).
At least in theory, the activities carried out by platforms may be classified as
intermediation between the aforementioned contracting parties. In addition to the main
service, ancillary activities may be supplied such as facilitation of payments, assistance in
the course of the relevant performance, internal dispute resolution mechanisms or
arbitration panels. It is tantamount to acknowledge that the fundamental feature at the
basis of the platform business model is the involvement of individual providers offering
services personally and mostly for the same platform (but not for the same client),
although being classified as autonomous workers. An arrangement as such has the merit
of lowering transaction costs (costs of finding the appropriate profile, negotiating an
agreement and enforce it), barriers to entry and information asymmetries – at least on
one side. Moreover, thanks to the reduction in agency costs, an “assetless company”
might arrange production inputs without incurring in the costs of formal property rights,
excluding those related to software, patents, and other intangible assets. That is why
tech companies are often able to scale very rapidly, thanks to a shift from asset control
to resource orchestration. In particular, such a way of arranging a digital infrastructure
results in the “pulverisation” of the stable employment relationship. Undeniably, it is
more convenient for clients and employer to engage workers on a task-by-task basis
rather than hiring them as employees. Many reports confirm that much of the advantage
of new platform- and app-based players derives from their failure to comply with labour
or social security regulations (
44
). On a closer inspection, this peculiar model allows the
(
42
) Communication from the Commission to the European Parliament, the Council, the European Economic and
Social Committee and the Committee of the Regions, A European Agenda for the Collaborative Economy,
COM/2016/0356 final (hereinafter “the Communication”).
(
43
) See T
ELLES
R. Jr. (2016), Digital Matching Firms: A New Definition in the “Sharing Economy” Space,
Economics and Statistics Administration Issue Brief, 1, available at: https://goo.gl/ZP9KgD.
(
44
) In order to be effective, this system implies the presence of a fungible and constantly available workforce.
See E
UROFOUND
and
I
NTERNATIONAL
L
ABOUR
O
FFICE
(2017), Working anytime, anywhere: The effects on the
15
management of a contingent workforce mobilised on a ‘just in time’ basis, thus
responding to demand peaks and shifting the impact of fluctuations on the worker’s
shoulders.
At first glance, platforms constitute a promising example of a two (or multi)-sided market
where participants are rapidly connected via desktop and/or mobile Internet. This
organisation has reversed the traditional ‘pipe’-based business model even though it
would be better to compare this new scheme to the network enterprise (
45
). On a closer
inspection, this model has existed for decades. What is new is the penetration of
(immaterial) infrastructures which determine frictionless transactions, not to mention the
“quantitative leap and exponential growth” in data and metrics that, collected, refined
and analysed, can “train” the internal algorithm thus resulting in an even more prompt
and successful matching function (
46
). The basic structure is the same in completely
different sectors, it replicates the original model of a virtual “bulletin board” such as
Craiglist or eBay, which are essentially advanced database (
47
). While a traditional firm
organises labour and other physical or immaterial resources minimising transaction costs
internally (according to the “make” vs. “buy” option, as theorised by Coase (
48
)),
platforms generate value by simplifying and supporting the interplay between providers
and users/consumers. Each successful interaction guarantees a 10 up to 20% transaction
fee to the benefit of the platform. At the same time, platforms have the possibility to
save fixed costs as well as to shed variable costs of production. Platforms do not incur in
high fixed costs, which results in large economies of scale. Accordingly, the only activities
that are “internalised” are middle management, sales and marketing functions. This is
how economies of scope can be combined with economies of scale and specialisation (
49
),
leading to a high-performing model of hierarchical outsourcing (
50
). Indeed, as argued,
“firms are able to make use of outsourcing without renouncing to hierarchy in the
management of the relevant business relationships by means of extra-legal mechanisms”
(
51
).
Contrary to what usually happens in value chain models, platforms make profits thanks
to the expansion of the ecosystem in a “circular and iterative” progression. Hence, a shift
occurs from mass production to large-scale networks, thanks to the involvement of
individual providers. As is well known, network effects increase proportionally with the
growing number of participants on the same side of the market (direct effects) or the
opposite side (indirect effects). There appear to be different pricing strategies for supply
world of work, Publications Office of the European Union, Luxembourg, and the International Labour Office,
Geneva.
(
45
) On the contrary, classical “pipeline businesses” operate by controlling a linear series of activities from the
inputs to the outputs (the value-chain model) and the process of production. Therefore, the main activity
consists in transforming inputs at one end of the chain into the finished product. See V
AN
A
LSTYNE
M.
W.,
P
ARKER
G.
G.
and
C
HOUDARY
S.
P. (2016), Pipelines, Platforms, and the New Rules of Strategy in Harvard
Business Review, 94(4), pp. 54-62.
(
46
) V
ALENDUC
G. and V
ENDRAMIN
P. (2017), Digitalisation, between disruption and evolution in Transfer:
European Review of Labour and Research, 23(2), p. 121-134.
(
47
) Early studies analysing the role of the Internet in enabling labour performances date back to 2001, see
A
UTOR
D.
H. (2001), Wiring the labor market in The Journal of Economic Perspectives, 15(1), p. 25
(demonstrating how the employer-employee matches would have been carried out on virtual venues rather
than on-site, making one’s location irrelevant on the global scale). See also T
ODOLÍ
-S
IGNES
A. (2017), The
‘gig economy’: employee, self-employed or the need for a special employment regulation? in Transfer:
European Review of Labour and Research, 23(2), p. 205 (describing platforms as “a new type of company
which claims to be a database”).
(
48
) C
OASE
R.
H. (1937), The nature of the firm in Economica, 4(16), pp. 386-405.
(
49
) G
OLZIO
L. (2005), L’evoluzione dei modelli organizzativi d’impresa in Dir. rel. ind., 2, p. 313. See also P
IORE
M. and S
ABEL
C. (1984), The second industrial divide, New York.
(
50
) As regard platform-specific investments, drivers must own a vehicle to be eligible to drive. For instance,
Uber usually assists its drivers in the relationship with an affiliated leasing company and sometimes
intervenes directly if the driver becomes delinquent in his payments. More recently, the company has also
designed a pilot insurance plan launched in partnership with Aon PLC to help cover on-the-job accidents.
See G
EE
K. (2017, August 8), In a Job Market This Good, Who Needs to Work in the Gig Economy?,
retrieved from http://on.wsj.com/2uBSChM.
(
51
) For a description of this phenomenon, admittedly in a different context, see D
E
S
TEFANO
V. (2009),
Smuggling-in flexibility: temporary work contracts and the “implicit threat” mechanism, ILO Working
Document, 4, p. 1.
16
and demand and asymmetric or surge pricing (
52
), not to mention the promotional
stratagems offering discounts and bonus for the launch of services (
53
). Conversely, the
end service (or, in the European Commission’s words, the underlying service – which is
the very service sold by the platform) is externalised to individual service providers. As a
consequence, also marginal costs of online platforms are low.
Yet, it could be said that “[n]o form of economic or social structure is ever entirely new
(
54
). Therefore, “Uberisation” does not redefine the notion of the firm. Rather this
phenomenon marks the shift from a bureaucratic control to a more sophisticated,
technocratic and invasive one. Hybrid forms adopted in the platform economy can be
situated formally halfway between those in market system (i.e. relationships driving “the
hardest possible bargain in the immediate exchange”) and those in network system
(characterised by “indebtedness and reliance over the long haul”) (
55
). On a closer
inspection, these relations may also be strongly shaped within the bureaucratic structure
of authority, in line with the hierarchical model (
56
). At the end of this scrutiny, it has
become clear that, while orthodox economic models of the firm, based on micro-analytic
transactional equilibria, are well placed to identify a general framework, they fail to
adequately describe the quite confusing and entangled reality of platform firms (
57
).
Accordingly, a fourth “breed” can be used to conceptualise the Uber-like firms’ prevalent
business model: a system combining complementary features belonging to dissimilar
prototypes but in different configurations.
2.1 An in-depth analysis on working conditions in three sub-
sectors: a multi-stage analysis.
Understanding the way the performance is carried out is pivotal, as in the light of the
principle of the “primacy of facts” explicitly enshrined in many national legal systems
substance has to be preferred over form (
58
). To explore working conditions on platforms,
secondary data from academic sources, non-structured interviews and reports has been
combined with publicly available information (“terms of service” as well as independent
investigations). By mapping information exchanged on closed-membership groups and
dedicated online forums, scrutinising workers’ conditions is one of the most effective
ways to collect information on legal regimes. The aim is to generate insights into a
representative portion of the growing multitudes of platform-facilitated arrangements by
accumulating disaggregated information about a range of different variables that help to
differentiate these working formats along various dimensions. In order to describe the
concrete operation of digital labour platforms, five key elements embedded in the
research question have been considered, focusing on their crucial legal implications in
terms of employment status classification: (i) access to the platform and registration, (ii)
selection process and hiring, (iii) command power and performance execution, (iv)
monitoring and rating (and deactivation), (v) payment rewards for completed tasks. A
(
52
) It seems that platform will be soon able to discriminate on the basis of the individual’s willingness to pay.
See E
ZRACHI
A. and S
TUCKE
M. E. (2015), Online Platforms and the EU Digital Single Market, University of
Tennessee Legal Studies Research Paper, No. 283, p. 9.
(
53
) This model turns customers into “ambassador” who are offered credits if they are able to successfully
convince friends to join the platform on both sides. See S
RNICEK
N. (2016), Platform capitalism, New York,
p. 46 (arguing that “[p]latforms often use cross-subsidisation: one arm of the firm reduces the price of a
service or good (even providing it for free), but another arm raises prices in order to make up for these
losses”).
(
54
) As noted by C
OHEN
J. E. (forthcoming), Law for the Platform Economy in U.C. Davis Law Review, p. 2.
(
55
) See P
OWELL
W. W. (1990), Neither market nor hierarchy: network form of organization in Research in
Organizational Behavior, 12, p. 302.
(
56
) For a complete analysis of these arrangements, see Part 3.
(
57
) G
RIMSHAW
D.,
M
ARCHINGTON
M.,
R
UBERY
J.
and
W
ILLMOTT
H. (2005), Introduction: Fragmenting Work Across
Organizational Boundaries in M
ARCHINGTON
M., G
RIMSHAW
D., R
UBERY
J. and W
ILMOTT
H. (Eds.), Fragmenting
work: Blurring organizational boundaries and disordering hierarchies, Oxford, p. 17 (arguing that “it is
more plausible to regard ‘market’, ‘hierarchy’, and ‘network’ as concepts that have proven valuable in
differentiating elements or dimensions of organising practices within and between organisations, rather
than as alternative designs of economic organisation”).
(
58
) See, for the Italian context, Cass. No. 13375 September 11, 2003, FI, 2003, I, 3321; Cass. No 4533 April
10, 2000, FI, 2000, I, 2196.
17
sort of chronological order will be followed in unfolding this multi-stage process, starting
with the registration and ending with the payment.
2.1.1 Access to the platform and registration.
To gain access to the “fleet” providing services through the platform, a worker needs a
smartphone equipped with high-speed connection. When registering, a prospective
worker has to provide biographical data (including age, location, highest degree, title,
grade, language skills, hobbies and know-how (
59
)) and a bank account where the money
will be conveyed. The weaker contracting party has only two options or, even better, a
“take-it-or-leave-it offer”: he or she may (i) adhere to the terms as drafted en bloc or (ii)
reject the clauses entirely. The worker cannot negotiate terms and conditions, as
electronic standard contracting forms – a kind of adhesion-styled agreement (
60
) – allows
only to click on “I Agree”. These long and complex forms contain a variety of clauses
encompassing several aspects of the relevant relationship, from forum selection to
dispute resolution, from worker classification to limitations of liability. Amidst such form
that many users might have not even read, workers can make inquiries and usually
receive template responses. These difficulties are likely to be worsened by the fact that
workers may lack direct channels of communication and are deprived of a collective
voice, thus leading to the impossibility to have an impact on the decision-making that
shapes labour processes.
One can identify a certain uniformity between platforms (
61
), and the unilateral standard-
setting activity may prove to be an effective way for platforms to auto-regulate the
relationship. Noticeably, the wording seems to be conceived in order to remain
ambiguous and prevent a negative outcome in case of lawsuit. In the European legal
system, this private standard-setting may also affect the assessment of the employment
status of workers. For instance, food delivery platforms usually require: being 18 or
older, an iPhone 4s or a superior version with a tariff scheme including a data connection,
willingness to work on the weekend, work permit, “sense of responsibility” (sic!) (
62
). In
addition to this, Uber demands documents to verify the eligibility (
63
). After logging in,
the worker is faced with a dashboard containing several elements (i.e. jobs completed,
assessment, profile or account, invoices). Speaking of requirements, in the peer-to-peer
transport sector, in most Member States the business model adopted by ridesharing
platforms is based on licenced drivers (
64
). In most of the cases, workers are defined
“partners” or customers, as they rely on the company’s service to be matched with final
customers. Moreover, workers renounce the possibility of challenging their status and
covenant to indemnify the platform in case of litigation (such clauses are far from
enforceable). Indeed, many companies now include arbitration agreements containing
waivers in their terms of service, thus narrowing the opportunity for employee
classification lawsuits (
65
). This results in the fact that crucial question issues related to
(
59
) Taking the German Clickworker as an example, worker can also upload and display “work samples”,
including of the sort of information that would have been considered outside the purview of the traditional
employer.
(
60
) Moreover, there are realistic reasons for arguing that both worker and costumers do not authentically agree
on anything when pushing the “I Agree” button, provided that these kinds of contracts require no explicit
manifestation of assent. See D
AVIS
N.
J.
(2007), Presumed assent: The judicial acceptance of clickwrap in
Berkeley Tech. L. J., 22(1), pp. 577-598.
(
61
) I
PEIROTIS
P. G. and H
ORTON
J. J. (2011), The need for standardization in crowdsourcing in Proceedings of the
workshop on crowdsourcing and human computation at CHI, p. 1.
(
62
) See https://rider.foodora.it/.
(
63
) Considering for-hire license, the list shall include documents relating to vehicle registration, certification of
commercial automobile liability, insurance policy.
(
64
) Some commentators have suggested that ridesharing services are flourishing in the UK due to the fact that
licensing requirements are typically minimal. Some member states (Ireland, France, and Spain) explicitly
impose that profits beyond the sharing of the expenses of a ride can only be made by licenced taxi or
private hire car drivers. For instance, when the application UberPop, which allows non-licensed drivers (i.e.
private citizens) to offer riders, was banned in Brussels, Uber restricted it offer to UberX and UberBlack,
only rely on licensed drivers, though this particular type of license is extremely easy to obtain.
(
65
) The issue pertaining potential conflicts of laws, particularly when transnational contractual parties are
operating in different jurisdictions, will be not addressed in this work. Broadly speaking, Rome I Regulation
18
legal classification are “decided secretly and lack precedential value” (
66
). After logging
in, the worker is faced with a dashboard containing several elements (i.e. jobs
completed, assessment, profile or account, invoices).
2.1.2 Selection process and hiring.
Work is allocated through open-calls rather than by direct assignment or sharp job role
requirements. The selection mechanism varies from platform to platform. Al least three
principal models can be described. The “hiring procedure” may be organised (i) as a bid
in response to a public auction (
67
) each worker must specify how fast and for what
price he or she could do the job (this often occurs mostly when the task is a creative
one), (ii) as an automatic matching operated by the internal algorithm on the basis of the
specification of the service required and the worker’s profiles, or (iii) in a different
approach, the worker may spontaneously apply for the fulfilment of the task. Some
“cognitive work” or “high-skill” platforms assist clients in launching a competition for the
services required. While many proposals may be submitted, only the “winner(s)” will be
paid when clients select the solution they prefer. More commonly, the client indicates the
project, workers submit their proposals and, if the solution is selected, they can arrange
the details, in few cases negotiating privately. A sense of freedom and flexibility is put
forth by assertions like the following: “[as] a Contributor, you and only you decide which
and how many tasks to complete, and when and where you complete them. You are free
to spend as much or as little time completing tasks as you choose. At no time are you
under any obligation to complete a task” (
68
).
Before getting a request, samples of work (together with a relevant portfolio, if any) can
be offered and assessed by experts. This preliminary stage is necessary to obtain the
possibility to participate in activities on many crowdsourcing platforms (on Upwork or
Clickworker (
69
), to name but a few). Similarly, workers aiming to use the German
platform Clickworker have to submit samples of their work or to undergo a test. They are
rated and subsequently offered tasks that match their score. This platform prevents
“unqualified clickworker” from accessing a set of task in the case they have “not yet
taken the respective assessments” or because the score falls below the threshold to be
reached. Reasons for exclusion include language requirements or skill-specific needs.
From a different point of view, it may prove to be hard to engage with a reliable client
due to the lack of information It should also be recalled that workers on challenge-based
platforms like Zooppa as well as those like Amazon Mechanical Turk (
70
) suffer the key
(Reg. 593/2008) set as a general rule the freedom of the parties to choose the applicable law (Art. 3).
However, it explicitly restricts this freedom through provisions laid down by Article 8 introducing the notion
of “objectively applicable labour law”. That is, the law that best suits to the relationship would be applicable
if no choice exists. The objectively applicable labour law has priority against the chosen law but only
regarding “provisions that cannot be derogated”, namely the hard part of labour law (ius cogens) and not
its flexible provisions that could be legally avoided by the parties (ius dispositivum). The Regulation offers a
set of criteria aimed at determining the objectively applicable law. These criteria are hierarchical and the
foremost criterion is the habitual place of work (lex loci laboris – Art. 8(2)). The level of protection cannot
fall below that which would be provided in the absence of choice. Additional criteria are the place of work
engagement (Art. 8(3)) and, ultimately, the place where work is more closely connected (rule of Art. 8(4)).
As it has been rightly pointed out, the legal construction of the “objectively applicable labour law” concept
is the element that should be kept and applied at a platform-mediate labour regulation at an international
level. See B
REGIANNIS
F.,
B
RUURMIJN
W.
J.,
C
ALON
E.
and
O
RTEGA
M.
A.
D. (2017), Workers in the Gig
Economy, Identification of Practical Problems and Possible Solutions, Tilburg University Working Paper.
(
66
) B
ISOM
-R
APP
S. and C
OIQUAUD
U. (2017), The Role of the State towards the Grey Zone of Employment: Eyes
on Canada and the United States in Revue Interventions économiques in Papers in Political Economy, (58),
p. 11.
(
67
) Some platforms (such as Boblr) charge a publishing fee for the launch of each competition.
(
68
) Crowdflower.
(
69
) In this case, worker must submit mock-up of their work or pass a test. The appraisal attributes a score
according to which a worker will be provided with certain types of tasks.
(
70
) According to Casilli, AMT is “the most prominent and high-profile example of a platform for micro-tasks”.
The first “Mechanical Turk” was designed and implemented by the Hungarian nobleman Wolfgang von
Kempelen in the late 1760s. The platform’s name pays homage to the eighteenth century mechanical
wooden cabinet containing a chessboard, life-sized, adorned with a turbaned mannequin that could
compete against human players at the game of chess. A small-bodied chess player was hidden and moved
19
risk of investing time, coupled with the unlikelihood of winning and the harshness of
jumping from a task to another, sometimes even with little or no effectiveness (
71
). It is
worth noting that Uber does not disclose the customer’s destination until the request is
accepted. This “blind” mechanism prevents drivers from competing with one another for
passengers. At the same time, platforms such as AMT do not provideany metrics on the
hourly rate of the work posted (
72
), therefore workers have one choice only, to complete
some and calculate the rate in order to understand whether the task is worth doing or
not.
2.1.3 Command power and performance execution.
As already explained, the managerial powers potentially or concretely exerted by a
platform may vary according to the type of the performance. In general, the overriding
evidence confirms that peer-to-peer transport services and delivery platforms exercise
remarkable supervision, direction, and control over how the ride is completed. For
example, cycle couriers – who accomplish their duty personally (
73
) – are strongly
recommended to wear a commercial uniform, use their own vehicle (
74
), show up in a
defined hotspot, then log in the app from their smartphone and wait for the first order.
While delivering fresh meals from restaurant to the customer’s address, they have to
follow the “suggested route”, in order to carry out their duty as quickly as possible.
Anyway, they are required to interact with the company’s middle management through
internal channels of communication (chat or apps such as “Staffomatic”) (
75
). Workers
have to pay for all running expenses (petrol, insurance, taxes) and assumes all
responsibility should an accident occur.
To this extent, both the (human or algorithmic) management and the final costumer are
responsible for ensuring and assessing the adherence to the instructions given. In the
“on-demand household services” subsector, instead, performances such as cleaning,
home repair, furniture assembling, gardening, painting, handyman-like tasks can be
commonly executed with a certain degree of autonomy, although workers have to
conform to guidelines and instructions. In this case, the dependence experienced by
workers is both economic and organisational. Conversely, it is undeniable that workers
using platforms such as GoPillar or the globally widespread InnoCentive (
76
) enjoy a
higher degree of autonomy. On the contrary, platforms such as Amazon Mechanical Turk
and Crowdflower deliver a highly standardised service.
pawns from inside – so no technology at all, aside from mock cogs and clockwork machinery. The
allegorical name represents a telling metaphor for a hoax, i.e. a piece of technology that relies on human
energies and intelligence in the “black box”. See S
CHWARTZ
O. (2016, February 29), Humans pretending to
be computers pretending to be humans, retrieved from https://goo.gl/MCbNFE. See also M
ARVIT
M.
Z.
(2014,
February
5),
How Crowdworkers Became the Ghosts in the Digital Machine, retrieved from
https://goo.gl/rJRXmA.
(
71
) According to an ILO survey, workers may spend up to 18 minutes of unpaid work for every hour worked
and paid. B
ERG
J. (2016), Income Security in the On-Demand Economy: Findings and Policy Lessons from a
Survey of Crowdworkers in Comp. Lab. L. & Pol’y J., 37(3), p. 557.
(
72
) M
ILLAND
K. (2017), Slave to the keyboard: the broken promises of the gig economy in Transfer, European
Review of Labour and Research, 23(2), p. 230 (explaining how “even poorly paying work gets done just
through this process”).
(
73
) More recently, a new contract has been introduced allowing riders in specific areas to provide the service
through a third party. In the light of a “substitution clause”, couriers may appoint a third person to
complete the task in their behalf. Nevertheless, in day-to-day practice, it truly is very difficult to select a
replacement, thus representing a barrier to the recognition of the employment status. In fact, in a potential
lawsuit, the burden of prove that those rights did not exist or were not exercised would be on the worker.
(
74
) As for TaskRabbit, “some clients provide tools and closely supervise the work, whereas others will expect
taskers to bring their own cleaning products or use their vehicles for transport-related tasks”, see P
RASSL
J.
and R
ISAK
M. (2016), Uber, Taskrabbit, and Co.: Platforms as Employers – Rethinking the Legal Analysis of
Crowdwork in Comp. Lab. L. & Pol’y J., 37(3), p. 645 (using the platform to show how different employer
functions are exercised by more than one entity).
(
75
) I
NVERSI
C.,
D
UNDON
T.
and
B
UCKLEY
L.-A. (2017), ‘I Don’t Work For, I Work With...’ Disposable workers and
working time regulation in the gig-economy, Paper prepared for presentation at the“5
th
Conference of the
Regulating for Decent Work Network” at the International Labour Office, Geneva, Switzerland, 3-5 July
2017.
(
76
) It enables research labs “broadcast scientific problems” with contests for prizes for solutions in several
fields.
20
This “faceless”, and sometimes “nameless”, employer may sanction behaviours that are
not considered compliant with standards or, even worse, are merely valued as lagging
behind customer expectations. In some cases, a training phase may be carried out either
before starting to work or should the rating fall below the critical threshold. Video-tutorial
are commonly used to promote certain best practices. As far as Uber is concerned, the
driver’s booklet (also referred to as “Community Guidelines”) invites to wear formal
attire, suggests to keep low the volume of the radio or to play classy music and
recommends offering chewing gums or beverages (
77
). Interestingly enough, some
platforms prevent crowdworkers from subcontracting the relevant task (or a fraction of
it) to other workers. This factor may be used to prove the lack of autonomy of the
contractual relationship. The irony is that Amazon Mechanical Turk, the biggest “human
cloud”
platform, imposes not to “use robots, scripts or other automated methods to
complete the Services”. Similarly, Crowdflower does not allow to employ “Internet bots,
web robots, bots, scripts, or any other form of artificial intelligence” (
78
) while
Clickworker prohibits outsourcing (“Clickworkers are expressly prohibited from
subcontracting or outsourcing projects to third parties unless this is expressly permitted
by the terms of a project description”) (
79
).
Some platforms are organising their workforce in shifts in a way contradicting the
message about the extreme schedule flexibility. Worse still, schedules are stable from
week to week (
80
) thus hampering workers’ elasticity and ignoring their needs in terms of
work-life balance. The narrative on the freedom of arranging one’s own timetable has to
be firmly demystified for two main reasons. Firstly, while it is correct to argue that
workers are free to be online whenever they want, the number of hours they spend
online is decisive when it comes to compensation. Moreover, the rates of rides accepted
and rides cancelled contribute to define the personal ranking. Lastly, when a system of
rotation is applied, the unpredictability of working hours is very problematic and
frustrating as it determines long periods of inactivity, sometimes due to miscalculations
of the potential demand and to arbitrary, dysfunctional day-to-day scheduling practices
(
81
). Despite the differences in the intensity of “temporal sovereignty” (i.e. autonomy), it
is evident that, in each of the reviewed sectors, workers must deliver a standardised
service. Perhaps most important, this reality contradicts the claim according to which gig-
workers should be considered entrepreneurs, enjoying the “freedom from formal
constraints such as mandatory working hours” (
82
). Moreover, ample evidence of the far-
reaching negative impact of the “permanent reachability” definitely holds true for
crowdworkers. Their constant “stand-by modus” is boosted by platforms using incentives
to compel then to be “logged-in” when needed, regardless their actual availability.
2.1.4 Monitoring and rating (and deactivation).
Traceability of data (
83
) is of utmost importance when it comes to vetting workers and
profiling customers (their needs, their locations and more importantly their
(
77
) O’Connor v. Uber Technologies, Inc., No C-13-3826 EMC, 2015. Moreover it imposes “not to shout, swear
or slam the car door”.
(
78
) See https://www.crowdflower.com/legal/.
(
79
) See https://workplace.clickworker.com/en/agreements/10123.
(
80
) Riders argue that their flexibility declined dramatically since the company has adopted a model based on
weekly shifts. A leaflet containing Deliveroo workers’ complaints and claims can be found at
https://goo.gl/Y1jPJt (in Italian).
(
81
) To make things worse, the “open availability” pattern makes it almost impossible to take on second jobs or
enrolling in training programs, and thereby contradicts the idea that these schemes may be used to
supplement income as a second job. The result may be a “trap”.
(
82
) L
EHDONVIRTA
V. (forthcoming), Work alerts and personal bests: Managing time in the online gig economy,
New Technology, Work & Employment.
(
83
) It is worth emphasising that the European Parliament resolution of 15 June 2017 on a European Agenda for
the collaborative economy (2017/2003(INI)) “[u]nderlines the importance of collaborative platform workers
being able to benefit from the portability of ratings and reviews, which constitute their digital market value,
and the importance of facilitating the transferability and accumulation of ratings and reviews across
different platforms while respecting rules on data protection and the privacy of all parties involved”. In
addition to this, the International Transport Forum, an intergovernmental organisation at the OECD, has
proposed that data collected from ridesharing platforms on the service performance of individual drivers
21
willingness to pay) (
84
). This issue has nevertheless been particularly salient in the
preliminary stage of on-demand economy’s development: platforms had been greeted as
an enhanced system allowing trust between strangers.
Trust, in turn, is based on a sort
of “digital market value”, whose very building block is the rating assigned to workers.
Therefore, it can be said that customers trust the fact that the worker’s career (or
professionalism) has been regularly traced and he or she can be considered accountable.
Several are the tools used to supervise the worker: GPS features on smartphones allow
delivery companies to constantly monitor the performance of their couriers, the five-star
system guarantees a constant appraisal of driver’s conducts, the internal software used
to complete click task is inherently meant to track actions. Platforms have essentially
deputized their customers to administer the workforce and make meticulous reports on
how service is rendered (
85
). Any provider’s act is therefore “transparent” to the eye of
the platform, not to mention the possibility of privacy violation, since the workers are
obliged “by design” to disclose personal information without a guarantee of
confidentiality. The same can be said for the client: sensitive information can be shared
with no clarity on their use. To put it bluntly, it can be concluded that technology-
mediated channels allow penetrating quality control, sometimes more effective than the
one exercised by a traditional employer.
As far as on-demand professional services are conceived, providers may take screenshots
of the provider’s monitor or track their keyboard strokes thanks to software specialized in
real-time surveillance, also assigning “control” tasks to double-check the quality of the
service (
86
). Even the amount of time a crowdworker has to complete a task can be set
and monitored: “[a] countdown timer displays the “Remaining Time” on each work page.
You must complete the job before the time runs out, i.e. reaches zero. If this happens
the job will be cancelled and will be lost for you[;] skipping a job ensures that you will
not be offered this individual job again, at least not within the next few minutes. As a
result, if few jobs only are available for you, you might not be offered any other jobs”.
Ride-hailing app, for instance, relay rides to another driver, if it is not accepted within 15
seconds. Platforms exchanging high-skill activities have developed a two-stage control
system. Before submitting the job to the client, a corrector checks, amends and
evaluates the proposal. In this case, the “current rating is an average of the last 25
evaluations [the worker] received for the jobs you submitted. When the corrector decides
whether the job goes back to [the worker] for reworking he has to include this evaluation
and a comment listing the improvements that need to be made”
(
87
). Changes in the
rating based on the comments can be followed in the “work history”. However, such
situation is temporary. A short time later, positive evaluations can increase your rating
while negative evaluations can decrease it.
On top of that, electronic ratings and reviews (defining the individual position in the
internal ranking) can grow up or fall down on the basis of the capricious customer
satisfaction systems, affecting the likelihood to be selected and hired in the future or to
be assigned the best paid tasks. Furthermore, the worker’s account can be deactivated
determining, in the most serious cases, the exclusion from the platform, which seriously
impairs his “career”. Unofficial sources refer that Uber driver’s account can be suspended
when his or her rating falls below 4,6 out of 5, though this condition can vary by city.
might be used by regulators, possibly as a partial substitute for licensing. See D
ARBÉRA
R. (2015), Principles
for the regulation of for-hire road passenger transportation services, OECD International Transport Forum.
(
84
) Data consists of two parts: firstly, data owned by the platform (e.g. the route determined through digital
maps) and, secondly, “data processed in the form of inferred information” (e.g. the locations of their
passengers’ work places and homes can be extracted from historical data). Data are being collected from
drivers and not from passengers and therefore can be mined even without the consent of passengers. See
Y
ARAGHI
N. and R
AVI
S. (2017), The Current and Future State of the Sharing Economy, Brookings Impact
Series, No. 03.
(
85
) R
OSENBLAT
A. and S
TARK
L. (2016), Uber’s Drivers: Information Asymmetries and Control in International
Journal Of Communication, 10(27), pp. 3758-3784.
(
86
) Software such as Worksnaps, Worksmart or Interguard sends regular screenshots from contractors’
computers so that clients can check and monitor the workflow or measure the time taken to complete a
task.
(
87
) See https://www.clickworker.com/faq/.
22
Ratings are averaged to mirror their last 500 trips. In particular, the platform places
drivers on notice of the kinds of conducts that may result in a negative consequence,
namely low ratings, high cancellations rates, low acceptance rates (below 90%). Al
already mentioned, the exclusion from the platform can be brutal or improvement
courses can be offered. In this case, a worker has to “provide proofs of the steps [he
has] taken to improve”. Understandably, rating system may determine an indirect lock-in
effect. As this portfolio cannot be transferred to another platform, a crowdworker is de
facto tied to a specific platform. Switching to a rival company is costly as it would mean
losing “the advantage of having good ratings” (
88
), thus entailing further concerns on the
ownership of this amount of data.
In addition to this, crowd-workers are faced with the volatility of customers’ preferences.
Currently, many online crowdworking operators’ terms of service allow unsatisfied
customers to retain a product without compensating the worker, with no explanation (
89
).
This could potentially lead to opportunistic or biased behaviours (
90
) or to an even more
condescending approach by workers, justified by the fact that they need to keep their
evaluation above the threshold (
91
). At the same time, should amendments be needed,
the worker must edit the work within a certain deadline, past which he or she loses the
right to be paid. The uncertainty is therefore exacerbated. Albeit incomplete about
consequences, the AMT clause is more than revealing: “If a Requester is not reasonably
satisfied with the Services, the Requester may reject the Services” (
92
). Ratings from
clients, as previously clarified, impact on the personal career and can determine whether
or not the worker receives further tasks. There is no appeal. Workers are rarely allowed
to challenge rejections they believe are groundless, unfair, or fraudulent. While it is true
that the evaluation is made after the execution of the performance, thus conflicting with
the idea of the constant provision of managerial power, it could be said that the implicit
threat of receiving a bad rating compresses the autonomy of the worker and governs his
conduct.
2.1.5 Payment rewards for completed tasks.
It can be said that the jobs created are insecure and extremely low-paid, for most
workers it is part of a piecemeal existence. In many cases, fares constantly fluctuate or
are altered without workers having any say or control, making the overall remuneration
hardly predictable (
93
). Special attention merits the severe income instability, as “the
system of remuneration and charging commissions often lacks clarity” (
94
). Low and
erratic incomes are justified by the fact that there is a pronounced oversupply of labour,
leading some workers to cut their rates below what they consider reasonable.
Two are the main models when it comes to put a price on a web-enabled work
performance: either the platform or the worker can set the payment rate, while in
almost all cases the platform handles the payments. Moreover, it could be said that
(
88
) W
AAS
B. (2017), Crowdwork in Germany in W
AAS
B., L
IEBMAN
W. B., L
YUBARSKY
A. and K
EZUKA
K., Crowdwork
– A Comparative Law Perspective, Frankfurt, p. 154.
(
89
) F
ELSTINER
A. (2011), Working the Crowd: Employment and Labor Law in the Crowdsourcing Industry in
Berkeley J. Emp. & Lab. L., 32, p. 153.
(
90
) See S
ILBERMAN
M. S. and I
RANI
L. (2016), Operating an Employer Reputation System: Lessons from
Turkopticon, 2008-2015 in Comp. Lab. L. & Pol’y J., 37(3), p. 505.
(
91
) The problem lies in the fact that algorithms might end up perpetuating human biases. See C
HERRY
M. A.
(2016), People Analytics and Invisible Labor in St. Louis U. L. J., 61, p. 12 (explaining how “if customer
ratings are as vulnerable to bias as research suggests, it is likely that minority drivers will be more likely
than white drivers to be deactivated, but the deactivation itself looks like an automatic event, divorced
from a person with bias”). For a comprehensive review on “workforce analytics”, see K
IM
P.
T. (2017),
Data-driven discrimination at work in Wm. & Mary L. Rev., 58, pp. 857-936.
(
92
) See 3.A and 3.B.
(
93
) One of the latest examples is an application introduced by UberEats, which has cut the “trip reward” from
initially £5 to £4 for weekday lunch and weekend dinner times, and to £3 for weekday dinner and weekend
lunchtimes. See O’C
ONNOR
S. (2016, September 8), When your boss is an algorithm, retrieved from
https://on.ft.com/2cJHEUA.
(
94
) D
E
G
ROEN
W.
P.
and
M
ASELLI
I. (2016), The Impact of the Collaborative Economy on the Labour Market,
Centre for European Policy Studies, p. 10.
23
compensation on a piece-rate basis is the norm (by the job or even by the task, from
cent amounts to two-digit euro sums) (
95
). More rarely, performances are remunerated
on an hourly basis, rarely with a “minimum rate” (
96
). When work is charged by the hour,
the monitoring activity is even more invasive. Accordingly, a relationship can be
established between the qualitative nature of the work and the payment scheme. When
prices are set by the platform, routine tasks have a fixed or variable fee and the fare is
calculated on current market factors. The platform can sometimes use specific
estimators. Contrariwise, creative activities are organised as a contest or a “lottery”
where “even those who don’t become ‘winners’ have to transfer all their intellectual
property rights to the platform” (
97
). In very few cases, rates of pay are negotiated
between the individual worker and the client or workers can freely set and advertise a
charge for specific activities.
With regard to food-delivery apps, two different payment structures can be found: a
piece-based remuneration or an hourly based one. In the first case, understandably,
workers are pushed to complete as many deliveries as possible within an hour, in a way
increasing competition and diminishing the sense of solidarity among colleagues. In the
second case, on top of a fixed hourly compensation, bonuses related to the number of
completed deliveries can be assigned. Moreover, platforms such as Uber, which pays on a
per trip basis, cuts their transaction fee after passing a certain number of rides. This
incentive-based promotional system is aimed at encouraging drivers to stay online and
available. Worse still, platform-coordinated workers are usually not allowed to solicit or
accept tips: acceptance of gratuities cultural practices is inflexibly discouraged (
98
). In
order to circumvent such obstacles, workers and customers have started gaming the
platforms: after a preliminary contact, they tend to interact directly, bypassing the
intermediary.
Interestingly enough, Amazon Mechanical Turk pays out cash only to American and
Indian bank accounts, being these the two main national communities of workers using
such platform. More than ten years after it was started, if workers are based in a
different state, they get paid in the form of credits or Amazon.com gift cards to be used
on the e-commerce section of the same website. As a consequence, workers may only
buy physical item and have to pay delivery charges – a win-win situation for the
platform. On Clickworker, instead, when a certain amount is reached, money goes to the
bank or to a personal Paypal account. Most workers work from 10 to 20 hours a week,
earning about $ 80 a month. Some workers who use the platform “full time” can reach
20-40 hours a week, amounting to a salary of $ 200-750 a month. The average hourly
pay is 2 euros. “Highly ranked” workers can get up to 6-8 euros per hour. As a result,
stress and unpaid waiting time are the norm.
(
95
) In general, the per-delivery rate is understood as a dramatic pay drop. Certain reports indicate that many
crowdworkers are economically dependent on one or few platform companies they are simultaneously
active on. For example, for approximately 40% of crowdworkers their work on Amazon Mechanical Turk is a
full-time occupation. See M
ILLAND
K. (2016), Crowd Work: the Fury and the Fear in
F
OUNDATION FOR
E
UROPEAN
P
ROGRESSIVE
S
TUDIES
(FEPS), The Digital Economy and the Single Market, Brussels, Belgium, p.
84. See also E
UROFOUND
(2015), New forms of employment, Luxembourg, p. 115 (describing poor
conditions of crowd workers: “research has, for example, shown that 25% of the tasks offered at Amazon
Mechanical Turk are valued at €0.007, 70% offer €0.04 or less, and 90% pay less than €0.07. This is
equals an hourly rate of around €1.44”).
(
96
) In a most unusual way, the global free-lancing platform Upwork imposed a sort of “minimum rate” of 3
dollars per hour for all jobs, irrespective of the location of the performance. Adtriboo, a Spanish platform,
applies a minimum or even fixed price for specific tasks on the basis of market prices and assumed number
of hours spent for completing such a task. See F
INCK
M. (2017), Digital Regulation: Designing a
Supranational Legal Framework for the Platform Economy, LSE Legal Studies Working Paper No. 15
(exhibiting a set of examples of self-regulation).
(
97
) D
ÄUBLER
W. (2017), Challenges to Labour Law in P
ERULLI
A. (Ed.), L’idea di diritto del lavoro, oggi: In ricordo
di Giorgio Ghezzi, Padova, p. 485. In a different work, the author uses the expression “greyhound”, see
D
ÄUBLER
W. and K
LEBE
, T. (2016), Crowdwork: datore di lavoro in fuga? in Giorn. dir. lav. rel. ind., 3, p. 479
ff. (arguing that crowdwork represents the dawn of an unprecendent form of compensating subordinate
work: contest with prizes).
(
98
) See W
EINBERGER
M. (2017, June 20), Uber will finally let its drivers accept tips, retrieved from
https://read.bi/2rRAaQU.
24
Notwithstanding the above considerations, the levels of remuneration for virtual unskilled
tasks appear particularly low when converted into hourly rates and compared with
national averages for standard employment involving similar tasks. They also appear
insufficient to serve as a primary source of income. The scale of employment associated
with this type of work is currently limited. Apart from the risk of rejection, one of the
main causes of complaint is delays in receiving payments. Other grievances underlying
dissatisfaction mention the fact that payments are often lower than expected. Although
the final fee is calculated according to the estimated average processing time as well as
to the level of difficulty, or – for creative tasks – to the special skills required to fulfil the
job, hidden extras may cause an significant reduction in the amount due. So, just in
conclusion, several negative aspects need to be stressed: firstly, the total lack of
transparency of criteria applied to payments, secondly, the absence of the minimum
levels of payment (either hourly or per task) and, thirdly, the worrisome unpredictability
of how much a worker may earn using a given platform.
25
3 Overview of the legal frameworks applying to labour
platforms.
From a regulatory perspective, platform labour represents a “moving target” (
99
).
Building on the results of the wide-ranging investigation into concrete working conditions
in three selected subsectors (passenger transport services, professional crowdsourcing,
on-demand work at the client’s premises), the crucial question that now needs answering
is whether the existing legal frameworks provide adequate responses to the current
(digital) challenges. This dilemma encapsulates very well tensions underlying any
discussion on working in the platform economy: is it sufficient that common statutory
rules and standards are correctly applied or is there a need for new legislation? On this
matter, boosters and naysayers have re-ignited a heated debate on the relationship
between law and technological progress or, even better, on the ability of regulations to
keep pace with various forms of innovation. Yet until very recently “little systematic
effort has been made to establish how existing regulations should apply” (
100
). Without
taking position on the unnecessary dispute, the following paragraphs will map and
highlight national requirements to comply with. One thing is sure: the claim that existing
laws should be disapplied and new ones adopted to better support digital matching firms
interests is weak (
101
).
To begin with, in the fragmented European landscape, several indicators help to define
the dimension of work associated with the platform economy. The stable proportion of
self-employment in the EU (14.9% in 2015 (
102
)) may serve as a proxy to determine the
potential size of a contingent of workers ready to engage in this economic segment (a
possible “‘surplus population’ of workers”). At the same time, both temporary contracts
and self-employment grew, quite strongly in some Member States, between the mid-
1990s and 2007. There is more. Despite the relatively tiny proportion of this class of
workers, there was an increase in the share of self-employed people without employees
(solo self-employed workers) between 2002 and 2015, while the composition of this
group is changing – service sectors account for the largest portion of self-employed
labour force (
103
). Moreover, the percentage of workers taking up second jobs has
augmented gradually thus expanding the same pool of workers: the majority of
freelancers in OECD countries are “slashers” (workers with a portfolio of multiple
activities to earn a living), as their contract work supplements another part-time or full-
time position (
104
). In spite of the differences among States, all these trends appear to
pre-date slightly the advent of the collaborative economy; it is indeed a well-known fact
that most digitally enabled companies were founded only in the last five to ten years.
(
99
) G
ARBEN
S. (2017), Protecting Workers in the Online Platform Economy: An overview of regulatory and policy
developments in the EU, European Risk Observatory Discussion paper, p. 14.
(
100
) H
UWS
U. and J
OYCE
S. (2016), The economic and social situation of crowd workers and their legal status in
Europe, op. cit., p. 4.
(
101
) D
EAKIN
S. and M
ARKOU
C. (2017, September 25), London Uber ban: regulators are finally catching up with
technology, retrieved from https://goo.gl/fpgTMM.
(
102
) Defined as “the employment of employers, workers who work for themselves, members of producers’ co-
operatives, and unpaid family workers. Employed people are those aged 15 or over who report that they
have worked in gainful employment for at least one hour in the previous week or who had a job but were
absent from work during the reference week”. See https://data.oecd.org/emp/self-employment-
rate.htm#indicator-chart. According to Eurostat, in 2016, 30.6 million people aged between 15 and 64
were self-employed in the EU. They accounted for 14% of total employment, see https://goo.gl/yc7jCK.
See also E
UROFOUND
(2017), Exploring self-employment in the European Union, Publications Office of the
European Union, Luxembourg (warning that the figure may seem “to be in contrast with the current
discourse on the rise of non-standard employment and selfemployment”).
(
103
) Own-account workers (self-employed people without employees) amount to 10% of all employed, while
employers (self-employed persons with at least one employee) take a share of 4.5% of total employment
in Europe. Among the self-employed, one out of five could be classified as “dependent” in the light of their
self-perceived and objective economic situation. For a complete picture, see E
UROFOUND
(2017), Non-
standard forms of employment: Recent trends and future prospects, Dublin, p. 1.
(
104
) H
USSENOT
A. (2017, August 21), Why the future of work could lie in freelancing, retrieved from
http://wef.ch/2vWWHPR. In addition to this, it is important to note that most of these self-employed
workers have no employees.
26
To put it bluntly, the very questions to be answered are: does an authentic loophole
exist? How can this space of uncertainty be reduced to the benefit of the parties
involved? Solving this dilemma is no minor matter. More recently, the European
Parliament has called on the Commission “to publish guidelines on how Union law applies
to the various types of platform business models in order, where necessary, to fill
regulatory gaps in the area of employment and social security”. At first glance, it might
seem that policymakers seem to have adopted a “wait–and–see” attitude so far, at least
from a labour law perspective: this approach is to be welcomed as a wise choice, given
the embryonic phase of this sector. Nevertheless, a number of episodic initiatives may be
traced mostly concerning work on demand and transport services (
105
). Conversely, much
more has been said (and done) in the field of competition, tax and administrative law,
particularly as regarding ride-haling services such as Uber or, more in general, occasional
services exchanged directly or through various intermediaries, including digital platforms
(
106
).
The truth is that some old issues gain new attention and conceivably importance in the
age of digital transformation of work. It can be said that, from a legal point of view,
“digital platform work” does not even exist. First of all, as explained in the previous
sections, the modalities in which digitally enabled work is performed differ much from
one another. While focusing on the “distribution channel” is convenient for academic
purposes (even though it is not a legal criterion for distinction), regulating app-mediated
work as an entirely unique category deserving “new notions, new concepts, new
regulations” should be avoided: the design of specific legislation only targeting platforms
makes little sense, at least in the area of labour law. Before going into the content of the
third section of the report, it is important to clarify that there has been a little number of
court cases around Europe providing insights on different topics yet. Thus this section will
highlight national requirements and legal frameworks that may be adapted, using the
already defined tripartite scheme to present collaborative economy services (peer-to-
peer transport services, crowdsourcing, manual on-demand work via platform).
Afterwards, the second paragraph will investigate the issues at hand in pending and
potential lawsuits and their possible outcomes, largely looking into the vexed question of
worker classification.
3.1 (European and) national requirements to comply with.
It should be made clear that, from a purely legal point of view, “digital platform-enabled
labour” does not even exist, in the sense that it is not “a sort of watertight dimension of
the economy and the labour market” (
107
). First of all, as explained in the previous
sections, digitally enabled work patterns differ much from one another. What is more, the
fact is that this intrinsic heterogeneity makes generalisation hazardous. While the wide-
ranging description of working conditions in the “app economy” is becoming consistent
and unambiguous at least in academic research, there is no consensus on the legal
(
105
) L
ENAERTS
K.,
B
EBLAVÝ
M.
and K
ILHOFFER
Z. (2017), Government Responses to the Platform Economy: Where
do we stand?, CEPS Policy Insights No. 30, p. 5 (claiming that “[i]n Europe, the most common approach
taken by governments involves applying the legal, regulatory and policy frameworks that are already in
place to the platform economy. In none of the countries studied is there a specific framework or guidelines
covering the platform economy as such”).
(
106
) Some countries such as Finland have published guidance on the application of the national tax regime to
the collaborative economy. Moreover, the Italian “Sharing Economy (Tax) Act” – which was simply brushed
aside when it came to the actual discussion introduced a “discounted” tax rate, with lower or no tax
charged on income up to €10,000. See V
AUGHAN
R. and D
AVERIO
R.
(2016), Assessing the size and presence
of the collaborative economy in Europe. PwC UK, impulse paper for the European Commission. Two other
significant experiences can be mentioned. For instance, in Belgium an act adds a new type of income
related labour activities delivered via “registered platforms” (i.e. recognised by the Belgian federal
government) below € 5,000 of annual income. A beneficial fiscal regime of 10% applies. Below this cut-off,
no VAT or social contributions are charged. For income gained through accommodation, there are different
rules (and then generally municipal taxes apply). In France, instead, in its 2014 annual digital report, the
Conseil d’Etat proposed the creation of a platform-specific status requiring the renegotiation of Directive
2000/31/EC.
(
107
) D
E
S
TEFANO
V. (2016), The rise of the “just-in-time workforce”: On-demand work, crowdwork, and labor
protection in the “gig economy” in Comp. Lab. L. & Pol’y J., 37(3), p. 472.
27
recipe for dealing with these issues. Scholars, commentators and lawmakers have
conflicting views on the applicability of existing legal frameworks to digitally enabled
platforms (
108
). On the one hand, those who think that new realities of work have
outgrown old-fashioned legal concepts, on the other, those who insist in making clear
that the newness of platforms does not deserve or justify a special treatment. Boosters
and naysayers have re-ignited a heated debate on the relationship between law and
technological progress or, even better, on the ability of regulations to keep pace with
transformational new realities. In the meantime, platforms have grown pretending not to
be covered by already existing regulations (
109
), “their preferred tactic being that of the
fait accompli (
110
).
Only few European States have adopted explicit regulations to address the numerous
issues stemming from the advent of the collaborative economy: the model is mercurial in
nature and a hefty intervention may provoke its premature asphyxiation. Accordingly,
before proposing universal or rather horizontal regulatory schemes, it would be useful to
validate the appropriateness of existing labour legislations. According to the European
Parliament, in fact, certain parts of the collaborative economy are covered by regulation
at local and national level” (
111
), therefore member States are encouraged to “step up
enforcement of existing legislation” by recurring to all available tools (
112
). In the view of
many scholars, “[m]ethods need to be found of applying existing European directives and
national legislation to work of this kind” (
113
). Taking note of the current state of play,
which kind of rule could be applied to digital labour platforms? As already mentioned, it
might be acknowledged that these arrangements often fall within the “extended family”
of “non-standard forms of employment” which are even more frequently used to “enable
a firm to adjust to fluctuations in demand” as well as for cost-saving reasons (
114
).
Typically, these forms imply a discontinuous and multilateral nature of the relationship
and the concrete risk of work intensification. Conceived as a balanced instrument
combining flexibility and security, they have largely been employed for unintended
purposes, thus resulting in an escalation of precariousness leading to a detrimental race
to the bottom.
If, for instance, platform-based labour is understood as a peculiar form of vulnerable (yet
subordinate) work, then the feasibility of the application of the three directives regulating
“atypical employment” to platform workers (namely, Part-Time Work (
115
), Fixed-Term
(
108
) S
UNDARARAJAN
A. (2017), The Collaborative Economy: Socioeconomic, Regulatory and Policy Issues,
Directorate General for Internal Policies Policy Department, Economic And Scientific Policy, p. 7 (arguing
that “it is important to think beyond simply trying to retrofit old regulatory regimes onto the new models”).
(
109
) A Uber manager once admitted that the company, today in the seventh year of operation, didn't “have the
time to marinate on some of the basics as the focus has been on launching in more cities”. See Uber
drivers confront challenges working for a ‘faceless boss’ (2017, June 9), retrieved from
https://goo.gl/WUFjK4.
(
110
) D
EGRYSE
C. (2016), Digitalisation of the economy and its impact on labour markets, WP ETUI, No 2, pp. 15
and 35 (claiming that “these […] strategies are likely to have a major impact on the traditional payment
and regulation models directly affecting workers”). See also G
RAHAM
M., H
JORTH
I. and L
EHDONVIRTA
V.
(2017), Digital labour and development: impacts of global digital labour platforms and the gig economy on
worker livelihoods in Transfer, European Review of Labour and Research, 23(2), p. 153 (arguing that “work
tends to be performed outside of the purview of national governments: minimum wages, worker
protections, and even taxes, seem to be optional rather than required for both the platforms and the clients
that source work through them”).
(
111
) European Parliament resolution of 15 June 2017 on a European Agenda for the collaborative economy
(2017/2003(INI)). Para. 14. Yet, it must be acknowledged that, in the section “Impact on labour market
and workers’ right”, the Parliament demands, “in a proactive way”, the modernisation of existing legislation
and the reduction in regulatory gaps.
(
112
)European Parliament resolution of 15 June 2017 on a European Agenda for the collaborative economy
(2017/2003(INI)). Para. 15.
(
113
) V
ALENDUC
G. and V
ENDRAMIN
P.
(2016), Work in the digital economy: sorting the old from the new, WP ETUI,
No. 3, p. 42.
(
114
) I
NTERNATIONAL
L
ABOUR
O
FFICE
ILO (2016), Non-standard employment around the world: Understanding
challenges, shaping prospects, Geneva, p. 159.
(
115
) Council Directive 97/81/EC of 15 December 1997 concerning the Framework Agreement on Part-time work
concluded by UNICE, CEEP and the ETUC, OJ [1999] L14/9 (pursuing quality alongside the non-
discrimination aims, the directive imposes comparable working conditions for full-time and part-time
workers unless there is an objective justification for dissimilar treatment). For a comprehensive
28
Contract (
116
), Temporary Agency Work (
117
)) should be assessed (
118
). Broadly speaking,
“each Directive reflects a trade-off between flexibility for employers to have recourse to
these forms of contract, coupled with standards on the treatment of such workers” (
119
).
They were underpinned by the assumption workers engaged in non-standard
employment should not see their working conditions limited by barriers erected on the
basis of the form of employment. Interestingly enough, the aforementioned directives
impose a prohibition of discrimination but also contain limits and provisions aimed at
preventing the abuse of such forms (
120
). Furthermore, one could claim the significance
of standards regarding Undeclared Work (
121
), Equal Pay (
122
) and Equal Treatment (
123
).
It can be said that the rationales behind these guidelines are now challenged by the
increasing prevalence of ITC-enabled arrangements, thus needing to be updated and
reviewed.
Another question to be addressed is whether, and to what extent, the Directive on
Working Time (
124
) offering a lot of flexibility in exchange for certain limits may be
used as regards the workers concerned. However, it has to be noted that workers who
can determine the organisation of their own working time are not necessarily excluded
from being in an employed relationship. Against this background, it has to be pointed out
that the notion of working time is defined as an interval “during which the person gives
assessment, see B
ELL
M. (2011), Achieving the Objectives of the Part-Time Work Directive? Revisiting the
PartTime Workers Regulations in Industrial Law Journal, 40(3), pp. 254-279.
(
116
) Council Directive 1999/70/EC of 28 June 1999 concerning the Framework Agreement on Fixed-term work
concluded by ETUC, UNICE and CEEP, OJ [1999] L175/43 (the directive establishes a framework for the
prevention of both abuse arising from the use of successive fixed-term contracts and discrimination against
those working under fixed-term contracts). For a comparative analysis of the origins and evolution of legal
regulations on fixed-term employment contract, see L
UDERA
-R
USZEL
A. (2015), Typical or Atypical:
Reflections on the Atypical Forms of Employment Illustrated with the Example of a Fixed-Term Employment
Contract-A Comparative Study of Selected European Countries in Comp. Lab. L. & Pol’y J., 37, pp. 407-
445.
(
117
) Directive 2008/14/EC of the European Parliament and the Council of 19 November 2008 on temporary
agency work OJ [2008] L327/9 (allowing workers to fall under the scope of domestic definitions of
“employees” or “workers” and qualify for stronger employment rights). See also E
UROPEAN
C
OMMISSION
(2014), Report from the Commission to the European Parliament, the Council, the European Economic and
Social Committee and the Committee of the Regions on the application of directive 2008/104/EC on
temporary agency work. COM(2014) 176 final, p. 19 (stating that the use of derogations to the principle of
equal treatment could, in specific cases, “have led to a situation where the application of the Directive has
no real effects upon the improvement of the protection of temporary agency workers”).
(
118
) The resolution of European Parliament urges the Commission to reinforce the implementation of already
existing directives devoted to precarious employment in order to “the spread of socio-economic uncertainty
and the deterioration of working conditions for many workers”.
(
119
) B
ELL
M. (2012), Between flexicurity and fundamental social rights: the EU directives on atypical work in
European Law Review, 37, p. 36 (wondering whether there is a fundamental social right to equal treatment
for atypical workers). The principle of equal treatment contained in directive on Part-Time Work, Fixed-
Term Contract, and Temporary Agency Work.
(
120
) See Paragraph 5 of the Framework Agreement on Fixed-Term Work; art. 4(1) of Directive 2008/104/EC.
(
121
) A type of undeclared work is “own account” or “self-employed work”, where self-employed persons provide
services either to a formal enterprise or to other clients, such as households.
(
122
) Council Directive 75/117/EEC of 10 February 1975 on the approximation of the laws of the Member States
relating to the application of the principle of equal pay for men and women.
(
123
) Council Directive 2000/78/EC of 27 November 2000 establishing a general framework for equal treatment
in employment and occupation.
(
124
)See R
ISAK
M. (2017), The scope of working time directive: perspective and challenges, A Thematic Working
Paper for The Annual Conference of the European Centre of Expertise (ECE) in the field of labour law,
employment and labour market policies: The Personal Scope of Labour Law in Times of Atypical
Employment and Digitalisation (arguing that “[those persons active in the virtual realms of the gig
economy] are formally free to work what and when they choose – but this freedom may often be no more
than formal due to an economic situation which does not leave them a lot of alternatives to selling their
labour in a certain way to certain contractual partners”). It has been claimed that the WTD has been
written “from the perspective of a standard worker is expected to work excessively long hours”. For a
comprehensive analysis, see D
AVIES
A. (2013), Regulating Atypical Work: Beyond Equality in C
OUNTOURIS
N.
and F
REEDLAND
M. (Eds.), Resocialising Europe in a Time of Crisis, Cambridge, UK, pp. 230–249.
Interestingly enough, AMT warns its providers to “comply with all applicable laws and registration
requirements, including those applicable to independent contractors and maximum working hours
regulations”.
29
up autonomy and is not free to determine its behaviour” (
125
), a concept that somehow
diverges from the basic structure of some of these alternative arrangements. At the
same time, one could argue that, in most of the cases, the substantial conduct and the
concrete situation of this group of workers approximates the one of employees, therefore
the provisions of the directive shall apply by analogy. Nevertheless, the extension of the
scope of the abovementioned regulation may be considered questionable, since “working
time restrictions in the sense to forbid [workers] to work beyond a certain number of
hours” (
126
) may prove be to the detriment of this group, infringing the possibility to earn
a decent wage, stay above the ranking thresholds, then be paid enough or be appointed
in the future. Indeed, tracking all activities in order to enforce a regulation as such
“would lead to a complete supervision of the individual and it would be incomplete
because the time spent, e.g., in preparing work with electronic media, will be not
included” (
127
). As shown above, an extensive interpretation of the Directive may cause
difficulties. In nearly all scenarios, its application is not entirely satisfactory, unless
prompt action is taken to ensure scrupulous revision and reasonable enforcement.
Box 1. The Communication on the European agenda for the collaborative
economy
Rebutting the charge of ill-adapting pre-existing or out-dated legal arsenal and national
regulations to this new mode of doing business (
128
), starting from 2015, the European
institutions have developed a framework for (collaborative) online platforms. This
journey’s milestones are several: after the adoption of the Single Market Strategy in
October 2015, the Commission focused its attention on carrying out a public consultation,
with the aim to gather the observations of various stakeholders including public
authorities, scholars, entrepreneurs and individuals, and a valuable Eurobarometer
survey (
129
). Finally, two Communications on Online Platforms and on the Collaborative
Economy were released between May and June 2016. Since the first is rather wide-
ranging (it contains principles such as a level playing field for comparable digital services,
responsible behaviour of online platforms, transparency, openness and non-
discrimination) (
130
), a hands-on analysis of the Agenda for the Collaborative Economy
will be provided (
131
).
(
125
) V
ERHULP
E. (2017), The Notion of ‘Employee’ in EU-Law and National Laws, A Thematic Working Paper for
The Annual Conference of the European Centre of Expertise (ECE) in the field of labour law, employment
and labour market policies: The Personal Scope of Labour Law in Times of Atypical Employment and
Digitalisation.
(
126
) R
ISAK
, ibidem.
(
127
) D
ÄUBLER
W., Challenges To Labour Law in P
ERULLI
A. (Ed.), L’idea di diritto del lavoro, oggi: In ricordo di
Giorgio Ghezzi, Padova, p. 497.
(
128
) H
ATZOPOULOS
V. and R
OMA
S. (2017), Caring for Sharing? The Collaborative Economy under EU Law in
Common Market Law Review, 54(1), p. 26.
(
129
) Flash Eurobarometer 438: The use of collaborative platforms, retrieved from
https://data.europa.eu/euodp/en/data/dataset/S2112_438_ENG
(
130
) Communication from the Commission to the European Parliament, the Council, the European Economic and
Social Committee and the Committee of the Regions, “A Digital Single Market Strategy for Europe”
{SWD(2015) 100 final}. Brussels, 6.5.2015. COM(2015) 192 final. See B
USCH
C., S
CHULTE
-N
ÖLKE
H.,
W
IEWIÓROWSKA
-D
OMAGALSKA
A. and Z
OLL
F. (2016), The Rise of the Platform Economy: A New Challenge for
EU Consumer Law? in Journal of European Consumer and Market Law, 1(6), p. 4 (soliciting the drafting of a
Platform Directive since “EU consumer law does not adequately cover the growing number of three-party
situations”).
(
131
) This paragraph draws from A
LOISI
A. (forthcoming), The role of European institutions in promoting decent
work in the collaborative economy in B
RUGLIERI
M. (Ed.), Multi-disciplinary design of sharing services,
Milano.
30
A cornerstone of the Communication resides in the classification of activities. If the
distinction between true and commercial sharing can easily be rooted in the switch
between “compensation costs vs. remuneration”, regulators are constantly asked to
distinguish between professionals and individuals who turn to the collaborative economy
platforms on an occasional basis, when services are provided for free or at a price that
barely covers costs. Experts suggest establishing a narrow set of criteria such as the
frequency with which a service is rendered, the provider’s profit-seeking motive and the
relevant payment (
132
). Yet, the issue is far from being unravelled. Lines between
categories are now increasingly tangled, and sometimes this uncertainty seems to be
sought deliberately in order to avoid due legal compliance.
In the Communication, the supply of the “underlying service” is considered the criterion
to establish which regulatory corpus should be applied to a given platform. In particular,
this paragraph focuses on the multilayer analysis designed by the Commission and aimed
at establishing whether a platform is operating in a “real world” sector, rather than as an
ICT matchmaker. According to the document, if the platform provides the “underlying”
service, in addition to the information society service one (
133
), it should be subject to
“relevant sector-specific regulation, including business authorization and licensing
requirements generally applied to service providers”. To this extent, the provision of the
“underlying service” has to be assessed in concrete by considering three key concurring
elements: (i) the price-fixing activity; (ii) the setting of other contractual terms; (iii) the
ownership of assets used to supply the service itself. Other criteria may be taken into
account: for example, the fact that the platform incurs in the cost and assumes all the
risk related to the service or whether an employment relationship exists with the worker.
The conditions are not extremely stringent as, for instance, merely assisting the
“ultimate” provider of the underlying service or arranging a ranking mechanism does not
“constitute proof of influence and control”. When most criteria are met, there are
vigorous indicators that the collaborative platform exercises a notable influence or control
over the ultimate provider. As a consequence, the platform cannot be merely considered
as the facilitator of the mere “intermediation service” (
134
), absolved of its sectorial
responsibilities. In Smorto’s words, as a first rule of thumb, when platforms exert a high
level of control and influence over workers, “they should be regarded as service
providers; conversely, when platforms limit their activity to the matching of demand and
supply, enabling peers to deliver the underlying services, they should be deemed as
intermediaries” (
135
).
(
132
) P
ETROPOULOS
G. (2017), An economic review of the collaborative economy, Bruegel Policy Contribution, No
5.
(
133
) See Article 1(2) of Directive 98/34/EC as amended by Directive 98/48/EC.
(
134
) Moreover, if the platform does not just act as a broker (even as a “notice board” or a “virtual display”), but
offers ancillary services, this cannot be read as a decisive index of influence or control over the underlying
service.
(
135
) S
MORTO
G. (forthcoming), Regulating (and self-regulating) the sharing economy in Europe: an overview in
B
RUGLIERI
M. (Ed.), Multi-disciplinary design of sharing services, Milano.
31
To move on to a different subject, it is crucial to scrutinise the analysis developed by the
Commission and aimed at establishing whether a worker is employed by the platform.
Equally noteworthy, the Communication refers to the notion of worker settled by the
Court of Justice (“…a person [who] for a certain period of time performs services for and
under the direction of another person in return for which he receives remuneration”).
Three criteria need to be met in order to detect the existence of an employment
relationship: (i) the existence of the subordination link; (ii) the performance of effective
duties; (iii) the presence of remuneration. The clarification of the Commission is relevant
to the extent of this analysis: the link of subordination can be described as the provision
of the directive power by the platform, which determines the content of the activity (i.e.
the scope of work), how the performance has to be accomplished and the form and level
of the remuneration. It has to be underlined that management and control on a
continuous basis is not decisive, as well as limited working hours or low rate of
productivity are not enough to exclude the existence of an employment relationship.
What is relevant is that an employment relationship can be identified, according to the
OECD’s assessment (
136
), when platform workers have no choice but to follow detailed
instructions given by the operator, or when the latter utilises customer ratings to control
or even “dismiss” providers. Taking a closer look, it is possible to connect this
examination with the previous one on the provision of the underlying service. If so, the
existence of an employment relationship would be sufficient to label the platform as a
provider of the off-line equivalent service, thus enforcing compliance with sector specific
requirements, in a circular way.
As for the nature of work and the presence of remuneration, the debate on how to set a
threshold for distinguishing between peers (or amateur/occasional providers) and
professional services providers is still intense. The European Commission support analysis
(
137
) mentions different elements: annual turnover for transport services, the frequency
of the activity (i.e. the services are offered regularly or marginally). The Parliamentary
resolution corroborates this view invoking “further guidelines on laying down effective
criteria for distinguishing between peers and professionals” (
138
). Taking a step
backwards, in France, for instance, an act was passed defining a “social liability” of
platforms towards (professional) self-employed workers. Due to this regulation, platforms
are obliged to pay insurance against industrial accidents at the workplace and
contributions to vocational training. In particular, a worker working for a certain amount
of time acquires rights in the field of lifelong training and the platform must pay the
contributions provided for by the legislation. Moreover, prior experience has to be
validated in order to be portable. Since August 2016, unprecedented provisions have
been introduced by the new Labour Code regarding “workers obtaining work through
digital platforms”. The Code extends to this group of freelancers individual and collective
rights, including to strike and form or join a union. In Ireland, last year, a bill was passed
allowing for self-employed people to collectively bargain (
139
).
The Communication concludes by advocating an intervention of Member States aimed at
“assessing the adequacy of national employment legislation” in relation “to the different
needs of workers and self-employed individuals in the digital world as the innovative
nature of collaborative business model” and to “provide guidance on the applicability of
their national employment rules in light of labor patterns in the collaborative economy”
(
140
).
(
136
) Other arguments rely on the “economic dependence” criterion, applicable to some providers.
(
137
) Commission Staff Working Document, Communication from the Commission to the European Parliament,
the Council, the European Economic and Social Committee and the Committee of the Regions – a European
Agenda for the Collaborative Economy.
(
138
) Paragraph 18. Paragraph 19, instead, demands “appropriate dividing lines”.
(
139
) O’R
EGAN
M. (2016, November 10), Seanad passes Bill providing collective pay bargaining for self-
employed, retrieved from https://goo.gl/whJaHP.
(
140
) C
AUFFMAN
C. and S
MITS
J. (2016), The sharing economy and the law: food for European lawyers in
Maastricht Journal of European and Comparative Law, 23(6), pp. 903-907.
32
3.1.1 Legal frameworks for casual work: on-call work, voucher-based,
and zero-hours contracts.
Platform-based working templates can be labelled under the umbrella definition of app-
driven casual arrangements, a subgroup of non-standard forms of work (
141
), rather than
under the self-employment category. In such arrangements “the relationships is not
stable and continuous, and the employer is not obliged to regularly provide the worker
with work, but has the flexibility of calling them in on demand”, in the words of
Eurofound (
142
). It has to be noted that the concrete performances rendered in the
context of crowdsourcing and work on demand via platform are collocated in a
problematic area of business where informality (implying recruitment by word of mouth
and cash payments), low wages and abuses are becoming increasingly frequent.
The “precarity” that originates from these arrangements is manifold. First, the very low
guaranteed hours mean that the worker can hardly earn sufficient income under the
contract. Second, while the intermittent (and unpredictable) nature of the hours provides
a flexible tool to the employer/client (
143
), the worker suffers from insecurity and
instability. Since the employer only agrees to pay for work completed, not to make work
available, the workers may have to wait for the next work order to materialise (what is
called “spot labour market”), and are therefore prevented from organising their schedule
ahead. Third, the fact that the employer can “adjust” the hours to the contractual
minimum, according to her needs, without any constrain, closely resembles “a situation
where the worker can be dismissed without notice” (
144
). The combination of these
vulnerabilities makes it “exceedingly difficult for workers to enforce the few rights that do
apply to them, as any challenge or inconvenience to the employer may lead to ad hoc
termination” (
145
).
Non-standard working arrangements could be seen as part of the broader trend of “over-
flexibilisation” of the labour market to the disadvantage of workers. This brief paragraph
will try to present work on demand as a species of the genus of casual work which has
been regulated in different ways throughout Europe. The common feature is the fact that
in this arrangement employer can determine “location and allocation of working time”
and the scope of the work to be done. In many cases, the performance is characterised
by “continuous employment relationship without continuous work” (
146
). Another shared
characteristic is that employers, who are supposed to bear the economic risk, end up in
shifting business risk on the workers. Despite the widespread factual complexity, work in
the on-demand economy can be characterised as a set of digitally intermediated
voucher-based contracts (
147
). Far from being monolithic, this tool represents rather an
“organising principle of precarious work”, kind of a wide and various arrangement where
(
141
) There is no universally accepted definition of NSFE. Note that Eurofound’s list contains only “new forms of
work” which is a narrower definition. E
UROFOUND
(2015), New forms of employment, Publications Office of
the European Union, Luxembourg (conducting a “mapping exercise” resulting in the identification of nine
distinct forms of employment as new – or of increasing importance since 2000). See also E
UROPEAN
C
OMMISSION
(2015), Employment and Social Developments in Europe 2014, Luxembourg; I
NTERNATIONAL
L
ABOUR
O
FFICE
(2016), Non-standard employment around the world: Understanding challenges, shaping
prospects, p. 2; OECD
(2016), New forms of work in the digital economy, Digital Economy Papers No. 260,
p. 17.
(
142
) The pattern is the same as for so-called “zero-hours contracts” that imply no minimum number of working
hours. It has to be admitted that the same report classifies a portion of digital labour platforms under the
label of ICT-enabled work. See E
UROFOUND
(2015), New forms of employment, Publications Office of the
European Union, Luxembourg, p. 46.
(
143
) E
ICHHORST
W. and M
ARX
P. (2015). Introduction: an occupational perspective on non-standard employment
in I
D
. (Eds.), Non-standard employment in post-industrial labour markets: an occupational perspective,
Cheltenham, UK and Northampton, MA, p. 5.
(
144
) G
ARBEN
S., K
ILPATRICK
C. and M
UIR
E. (2017), Towards a European Pillar of Social Rights: upgrading the EU
social acquis, College of Europe Policy Brief, 1, pp. 4-5.
(
145
) Ibid.
(
146
) As explained in V
ALENDUC
G. and V
ENDRAMIN
P.
(2016), Work in the digital economy: sorting the old from
the new, WP ETUI, No. 3, p. 34.
(
147
) F
REEDLAND
M. R. and P
RASSL
J. (2017), Employees, Workers, and the ‘Sharing Economy’: Changing
Practices and Changing Concepts in the United Kingdom, University of Oxford, Legal Research Paper Series,
No 19.
33
workers are not guaranteed a fixed (or even any) number of hours in a given period and
where, at least in theory, they are not obliged to accept any offers of work made by their
employer/principal/client (
148
). This controversial area needs to be addressed by
European authorities in order to improve the situation of these workers.
Member states have inconsistent regulations for casual work, the Dutch model (
149
) and
the British scheme (
150
), being on two opposite ends of a spectrum. This legal format can
be used to regulate arrangement in the work on demand via platform (at the household’s
premises) segment, but it may fail to regulate different activities. In The Netherlands, no
minimum hours are assigned and voucher-based workers are paid only if actually
worked. A minimum wage is guaranteed for oncall workers employed for less than 15
hours per week. Nevertheless, after three months, the worker (who has worked for a
certain amount of time) must be guaranteed a minimum working hours on the basis of
the average hours worked in the previous quarter; social partners may adapt, and in
some cases alter, certain modules for a specific sector in a collective agreement. The aim
of the law was to discourage the use of the oncall contract as a sham for what should be
a SER. According to commentators, the Dutch experiment was successful.
Conversely, in the UK, zero-hours contract (ZHC) workers are not entitled to any pay if
the employer cannot provide them with work. In this system, workers are not paid for
their inactive time away from the workplace, while they receive remuneration for the
time spent working or waiting at the employer’s premises. In the UK, ZHC are considered
as employment contracts only when: (i) an obligation to provide work personally is
applied; (ii) employer and employee share a “mutuality of obligation”; (iii) the worker
agrees to be under the “sufficient” control of the employer, either expressly or implicitly
(
151
). Nevertheless, it is very difficult “for an individual to acquire sufficient continuity of
employment to qualify for significant employment rights, including unfair dismissal and
redundancy protection which require two years of service” (
152
).Dissimilarly, the Irish
model can guarantee a certain degree of protection to workers, but such approach is not
very widespread (
153
). If a worker is not called at all in a given week, she is entitled to a
compensation for the availability to be engaged amounting to a fraction of the contract
hours (25% of the available hours, with a cap). Just like the platform labour
arrangements, ZHCs have no specific legal status.
In France, a voucher-based scheme (the so-called “chèque emploi service universel”) has
existed since 1994, then modified in 2006 as for domestic work childcare. A very similar
instrument was available in Italy until the beginning of 2017 (
154
): each “voucher” –
easily purchasable had a nominal value of 10 euros (7.50 for the worker and the
remaining part allocated to the national system of social security and for health and
(
148
) A
DAMS
A.,
F
REEDLAND
M.
R.
and
P
RASSL
J. (2015), The ‘Zero-Hours Contract’: Regulating Casual Work, or
Legitimating Precarity?, in Giornale di diritto del lavoro e di relazioni industriali, 147, p. 529
(
149
) Introduced by the Flexibility and Security Act (FSA) in 1999.
(
150
) A
DAMS
Z. and D
EAKIN
S. (2014), Re-regulating zero hours contracts. Institute for Employment Rights,
Liverpool.
(
151
) For a complete review of various tests, see B
URCHEL
L
B.,
D
EAKIN
S.
and H
ONEY
S.
(1999), The employment
status of individuals in non-standard employment, Department of Trade and Industry, London, p. 5 ff. and
also p. 11 for other factors taken into account by courts.
(
152
) D
EAKIN
S. (2016), New forms of employment: Implications for EU
law The law as it stands in W
AAS
B.
(Ed.) New forms of employment in Europe, Bulletin of comparative labour relations, 94, Alphen aan den
Rijn, The Netherlands, p. 49.
(
153
) For a complete overview, see O’S
ULLIVAN
M.,
T
URNER
T.,
L
AVELLE
J.,
M
AC
M
AHON
J.,
M
URPHY
C.,
R
YAN
L.,
G
UNNIGLE
P.
and O’B
RIEN
M.