Revista de economía mundial 66, 2024, 129-149
ISSN: 1576-0162
DOI: http://dx.doi.org/10.33776/rem.vi66.8065
Precariousness in emPloyment mediated
by digital Platforms. evidence from euroPe
Precariedad en el emPleo mediado Por
Plataformas digitales. evidencias de euroPa
María Isabel de Andrés
Universidad de Huelva
maribel.ap2000@gmail.com
Emilio Congregado
Universidad de Huelva
congregado@uhu.es
Concepción Román
Universidad de Huelva
concepcion.roman@dege.uhu.es
Recibido: noviembre 2023; aceptado: enero 2024
abstract
Is precarity inherent to employment when it is mediated by a digital
platform, or does employment precarity have other causes? Using the first wave
of the European survey on collaborative economy and employment (COLLEEM,
hereinafter), we identify different types of precarity among platform workers
by using different operationalizations of this phenomenon. Our results indicate
that i) the probability of precarity in on-demand platform work varies according
to the type of employment and to certain sociodemographic characteristics;
ii) findings are sensitive to the dimension of precarity that we address; and iii)
self-employed individuals and those workers who access digital platform jobs as
last resort have a more positive perception of working conditions in the sector
than salaried employees and those whose reason for entry was not the lack of
alternative employment. The study provides guidelines for the effective design of
mitigation policies to protect workers in the digitalized EU labour market.
Keywords: Platforms, gig economy, precarity, contingent work, self-
employment, underemployment.
resumen
¿Es la precariedad una característica inherente al empleo mediado por
plataformas digitales u obedece a otras causas? Utilizando la primera oleada
de la Encuesta Europea sobre Economía Colaborativa y Empleo (COLLEEM,
en adelante), este trabajo identifica los empleos precarios en plataformas
mediante diferentes operacionalizaciones de este fenómeno. Nuestros
resultados indican que i) la probabilidad de encontrarse en situación de
precariedad en el empleo bajo demanda varía según el tipo de empleo y ciertas
características sociodemográficas; ii) los hallazgos son sensibles a la dimensión
de precariedad que consideremos; y iii) los trabajadores autónomos y aquellos
que acceden a empleos en plataformas digitales como último recurso tienen
una percepción más positiva de las condiciones laborales en el sector que
los empleados asalariados y aquellos cuya razón de ingreso no fue la falta de
empleo alternativo. El estudio proporciona pautas para el diseño efectivo de
políticas de mitigación orientadas a proteger a los trabajadores en el mercado
laboral digitalizado de la UE.
Palabras clave: Plataformas; economía colaborativa; precariedad; trabajo
contingente; trabajo autónomo; subempleo.
JEL Classification / Clasificación JEL: J08 J23 J81 D20 O35.
Revista de economía mundial 66, 2024, 129-149
1. introduction
The chronification of underemployment after the Great Recession (Bell
and Blanchflower 2017, 2018; Borowczyk-Martins and Lalé 2020, Valleta
et al. 2020, Congregado et al. 2024) and the progressive substitution of
regular employer-employee relationships into contractual relationships
(Drahokoupil and Fabo 2018) are two of the most salient features of today’s
labour markets. Some scholars associate the upsurge of gig work with the
resurgence of precarious work1 (Abraham et al. 2017; Stanford 2017; Coyle
2017; Montgomery and Baglioni 2020; Kilhoffer et al. 2020), and others
argue that the penetration of platform-mediated labour has been particularly
intense in countries where employment protection legislation is more stringent
(Fabo et al. 2017; Congregado et al. 2019). From this perspective, employers
would have found, in on-demand digital work, a way to evade some elements
of labour legislation and devise more flexible forms of employment, including
dependent forms of self-employment (Muehlberger 2007; Eichhorst et al.
2013; Roman et al. 2013; Stewart and Stanford 2017; Graham et al. 2017).
As a result, the power imbalance in the relationship between platforms and
digital workers, which is inherent in the very architecture of platforms, has
called into question the effectiveness of labour legislation, while it is seen as
eroding social contracts and workers’ rights (De Stefano and Aloisi 2019).
However, we cannot ignore that platform-mediated employment has also
fostered the emergence of new forms of employment that cannot be linked
to precarity. These new forms include those professionals and freelancers
whom digital platforms have allowed to expand their client network or become
a hybrid entrepreneur, combining his/her salaried job with a second job as
a self-employed worker (Kenney and Zysman 2016; Malo 2018; Kässi and
Lehdonvirta 2018).
In the same way, we cannot ignore the fact that the perception of precarity
depends on the starting situation. For some platform workers, the decision to
participate is a voluntary decision, which they value more highly than those who
1 Although there is no consensus definition on the concept and dimensions of precariousness,
the conventional wisdom considers that the insufficiency of income and employment instability
are the two vectors that determine the precariousness of a job (Kalleberg, 2009, 2011; Olsthoorn,
2014). Orn this basis, with the term precariousness, in this study, we refer not only to the overlap of
uncertainty, instability and underemployment, but also to the development of forms of dependent
self-employment. Section 2.2. present a more detailed discussion about this concept.
132 María Isabel de Andrés · Emilio Congregado · Concepción Román
have previously held regular jobs. These platform workers include those who
have accessed platform-mediated employment out of necessity because their
low employability meant that they did not receive employment opportunities
in the conventional labour market and for whom platform employment became
the only way to leave unemployment or inactivity.
Thus, we would argue that previous research seems to have overlooked the
differences between different types of digital platform workers. Professionals,
hybrid entrepreneurs and people with low employability who decide to become
platform workers due to the lack of job opportunities in the traditional labour
market belong to a distinct group that could exhibit different perceptions
about the pros- and cons- of the on-demand work offered on digital platforms.
This heterogeneity suggests that precarity, in its different manifestations, may
not be a phenomenon that occurs simultaneously with the fact that the match
between provider and job seeker is carried out through a platform. Rather, it
may be associated with the prevalence of certain employment sectors and
activities that predominate on platforms2 and with the population groups with
lower employability that accept such on-demand jobs as a last resort, even
when these positions are unstable, hinder them from reaching certain salary
standards, or prevent them from having sufficient coverage by employment
protection legislation (Amuedo- Dorantes 2000; Vosko et al. 2010).
Research on the precarity in digital platform work should thus deepen
the identification of the existing heterogeneity mentioned above because
of two important reasons. It is important to combat practices that inhibit
the effects of labour legislation, and it is equally important that the policies
against precarious forms of employment on digital platforms do not become
an obstacle to the articulation of flexible forms of professional employment or
to access to the labour market for groups with lower labour participation rates
and greater risk of exclusion.
Contributing to a better understanding of precarity in the platform economy
seems crucial to improve the accuracy of the so-called mitigation policies, i.e.,
the effectiveness of those measures aimed at protecting this sensitive group of
workers. However, the number of related empirical studies has been rather low
to date, despite this topic being a hot political issue. The multifaceted nature of
precarity (Bell and Blanchard 2013, 2018; Borowczyk-Martins and Lalé 2020;
Valletta et al. 2020; Roman et al. 2011; Williams and Horodnic 2018; Pantea
2021), the heterogeneity of situations involved in digital platform work (Pesole
et al. 2018; Congregado et al. 2019, 2022; Brancati et al. 2020), the lack
of a satellite account for measuring the digital economy (Ahmad & Ribarsky,
2019) and, ultimately, the lack of reliable data are seen as probable causes of
this research gap (Huws et al. 2017, 2019; Pesole et al. 2018; Congregado et
al. 2019; Brancati et al. 2020). Furthermore, most studies focus on a single
2 Currently, platform-mediated employment has nearly monopolized some activities, e.g., “delivery”,
where temporary contracts, subcontracting, forms of dependent self-employment and fixed-term
contracts are common practices.
133
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
country or are based on semistructured interviews among stakeholders or
platform workers (e.g., Kilhoffer et al. 2020; Schor et al. 2020). However, to the
best of our knowledge, an empirical analysis that characterizes the precarity
in on-demand platform work in a comparable cross-country setting does not
exist to date.
Filling this research gap is precisely the main aim of this article, shedding
some new light on the precarity in platform work by using alternative
operationalizations of the underemployment concept. Our main hypothesis is
that platform-mediated employment is heterogeneous, and underemployment
is not intrinsic to it but rather is confined to certain types of employment where
the workforce is more abundant and less qualified.
In doing so, we perform a cross-country analysis of the precarity in on-
demand platform work with European data. Europe is a suitable case study
because European authorities have carried out direct interventions to prevent
the incorrect use of nonconventional working arrangements in digital platform
work to evade employment protection legislation. Furthermore, European
countries are those with more regulated labour markets and in which these
types of practices are more prevalent than in other countries where flexibility
and rotation are common features of traditional contractual relationships.
Finally, although it would be interesting to apply this type of comparative
analysis to a larger number of countries, as it would allow us to analyse the
effects of the penetration of this type of labour relation in different institutional
frameworks, there is no internationally comparable database other than the
one used in this paper—the COLLEEM—and the one constructed by Huws et
al. (2017, 2019). The advantages of the sampling design in COLLEEM make it
preferable for use to the database constructed by Huws et al. (Kilhoffer, 2021).
The present paper relates to two types of literature. Firstly, to the general
literature on the determinants of precarity. (Büchtemany and Quack 1990;
Kretsos and Livanos 2016). Secondly, to the literature on platform labour
(Drahokoupil and Piasna 2017; Malo 2018; Congregado et al. 2019; Bogliacino
et al. 2019).
In this context, our paper contributes to the emerging field within the
labour economics literature focusing on precarity in digital platforms, being
one of the few contributions examining precarity on digital work platforms with
empirical data. In contrast to fragmentary or partial approaches, here, precarity
is addressed as a multidimensional phenomenon, comprising job instability,
underemployment, the perception of protection, rights and bargaining power.
Some of these terms are alternative demarcation criteria defining precarity.
Finally, our results should also assess whether further EU action on platform
work is merited or whether some elements should be rethought.
The remainder of the article is structured as follows. In Section 2, we
conduct a brief literature review to establish the potential links between the
emergence of all atypical work models associated with digital labour platforms.
Chronic underemployment situations and the loss of workers’ autonomy and
rights are analysed as the main outcomes of platform work precarity. We also
134 María Isabel de Andrés · Emilio Congregado · Concepción Román
derive hypotheses regarding the integration of work and precarity into the
platform. We then test these hypotheses using the data from COLLEEM. After
justifying the use of this source of data, the variables that we employ from it
and the methods applied are discussed in Section 3. Section 4 describes the
empirical results, and finally, Section 5 summarizes themain conclusions and
limitations and provides an outlook for further developments.
2. related literature
In this section, we try to determine the intersections between the attempts
to characterize nonstandard forms of employment on digital platforms and to
discuss how the phenomenon of precarity has been defined in the previous
literature. Accordingly, this review provides us with the opportunity to
effectively frame our analysis.
2.1. Work mediated by digital Platforms
Employment mediated by digital platforms refers to the provision
of services where both employer–employee matching and payment are
carried out through a digital platform. Here, a digital platform is specifically
commissioned to mediate between the solicitor of a service and the platform
worker. In addition, it fulfils this role regardless of how a service is provided—
online, on location or mixed—or whether a salient employment activity is part-
time or occasional or if, in contrast, it constitutes an employee’s principal
activity. While this definition suits so-called crowd workers (Lepanjuuri et al.
2018), some researchers prefer to reserve the term crowd work for jobs that
are completely online or location independent (De Stefano 2016), using gig
work to refer to activities that are provided in person (Brancati et al. 2020).3
However, regardless of the tradable nature of a service, workers whose
services are mediated through digital platforms clearly constitute an extremely
heterogeneous group with variations in the following: their specializations and
required qualifications (De Groen et al. 2017); their motivations, e.g., flexibility,
work-life balance or lack of opportunities; their level of dedication or intensity
of employment (Pesole et al. 2018); or their particular work situation (Todoli-
Signes 2017).
These platforms have shaped a new unit of division of labour, i.e., the task,
which has contributed to the outsourcing of numerous processes to make
demand more flexible and to avoid signalling problems with productivity,
reducing many costs associated with the protection of regular employment
(Larsson and Teigland, 2020).
As a result, there has been a profound change in the labour relations
system and the generalization of certain forms of precarity. On the one hand,
3 For more on these terminological issues, see O’Farrell and Montagnier (2019); or Koutsimpogiorgos
et al. (2020), among others.
135
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
the generalization of false self-employment (Drahokoupil and Fabo 2016), the
replacement of traditional labour relations by atypical forms of employment
and the use of these to avoid rigidities and costs due to labour legislation
(Standford 2017) are at the centre of the debate on the value of these new
forms of (nonstandard) employment. On the other hand, digital platforms have
led to forms of precarity that are associated with uncertainty and job instability
due to the very nature of these jobs/tasks (Webster 2016; Kuhn and Maleki
2017; Lehdonvirta 2018; Berg and Rani 2018; Sutherland et al. 2019; Schor
et al. 2020).
In short, although the growing importance of platform-mediated
employment can be understood in a more general context of decreased
transaction costs for employers via the fragmentation of tasks and activities,
the disruption it produces changes the costs and risks for different actors in
the labour market. Thus, this process has increased the transaction costs for
workers, who have to acquire skills and invest time in their search process.
Furthermore, as the bulk of the employment mediated by these digital work
platforms has been associated with activities that require low skill levels,
these jobs have been carried out by less specialized groups of workers, whose
potential for substitution is greater. For these less-skilled workers, platform-
mediated employment has therefore entailed a loss of bargaining power and, to
a large extent, a loss of the protections that certain labour market institutions
grant to traditional forms of employment (Drahokoupil and Piasna 2017).
Accordingly, a solid characterization of the phenomenon of precarity in work
mediated by digital platforms is needed to guide the search for and design of
corrective instruments, given that such analyses are scarce, incomplete, and
limited to national contexts. This literature is reviewed in the following section.
2.2. the different dimensions of Precarity
Precarity is a multidimensional concept associated with jobs where
uncertainty, instability4 and other conditions place a worker in a situation
of vulnerability (Rodgers and Rodgers 1989; Kalleberg 2009). This includes
any manifestation of the phenomenon of underemployment (Feldman 1996),
situations in which a worker’s bargaining power or level of work protection is
below those that accompany a traditional job (García-Pérez et al. 2020) or
cases where salaried employment is disguised as (dependent) self-employment
(Roman et al. 2011).
Precarity is commonly associated with fixed-term contracts, involuntary
part-time employment or below-standard earnings (Olsthoorn 2014). However,
these are not the only dimensions explored. Precarity has also been associated
4 Job insecurity does not always mean precarity. However, the perception of the insecurity depends
on the expected duration of the spell of unemployment. From a European perspective, and in
particular, from countries with rigid labour markets, the perception of insecurity is usually associated
with precarious forms of work.
136 María Isabel de Andrés · Emilio Congregado · Concepción Román
with subjective perceptions of job satisfaction, poverty thresholds, vulnerability
and other workplace aspects, such as the extension of the workday or the
existence of controls and flexibility (Arranz et al. 2018). Therefore, we believe
that an adequate definition of the concept of precarity should establish its
complexity by defining it according to its different dimensions. This is the
approach we adopt in this work.
Thus, our first approach to the phenomenon entails identifying precarity
with the different forms of underemployment. Labour statistics characterize
underemployment as work that involves involuntarily performing a job with
a shorter duration than normal and searching for or being willing to accept
more hours or another job until a workday is complete (Hussmanns et al.
1990). This definition, however, omits some forms of underemployment. Thus,
a broader definition of underemployment encompasses work that involves, at
least i) performing, involuntarily, a type of work that requires a lower level of
qualification, skills or experience than those available to a worker (Feldman
1996); ii) work that is not adjusted to the professional or study field of a
worker (Feldman 1996); iii) involuntary, part-time, temporary or discontinuous
work (Findeis et al. 2000); iv) earnings that are below the standard or below
80% of a worker’s pay in his or her previous employment (Findeis et al. 2000);
and v) sufficiently low levels of worker satisfaction to leave a worker open to job
changes (Bell and Blanchflower 2019).
Hence, the literature on underemployment has focused on analysing its
determinants and countercyclical nature. Its findings clarify that there is a
certain concentration of underemployment among population groups with
lower employability. Thus, precarity seems to have a greater prevalence among
women (Hakim 1997; Wilkins 2006; Jefferson and Preston 2010; Weststar
2011), elderly individuals (Bell and Blanchflower 2013), and young people
without experience or with a low educational level (Wilkins 2006). However,
the results concerning the pure effect of educational level on the probability
of being underemployed are not conclusive (Wilkins 2006; Weststar 2009).
With respect to job attributes, the literature indicates that underemployment
is more prevalent in sales and service work, in sectors where there are higher
proportions of part-time and temporary jobs (Weststar 2011; Wilkins et al.
Wooden 2011; Tam 2010) and in professional categories that do not require
high levels of qualification (Tam 2010). Empirical evidence has also linked the
incidence of underemployment with self-employment (Bell and Blanchflower
2013 and Wilkins 2006), a link that is reinforced by the notion that some
people use self-employment as a last resort amid job scarcity or a lack of
other job opportunities (Moore and Mueller. 2002). Finally, regarding working
life, the literature suggests that workers who have been laid off, who have
long been unemployed or who are stuck in their career are more likely to be
underemployed (Feldman 1996; Wilkins 2006).
As we noted above, many analyses have focused on underemployment
behaviour during different phases of the employment cycle. Therefore, it seems
that the anticyclical nature of underemployment largely reflects company
137
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
responses to shocks in demand (at least those of a temporary nature, such as
those during the COVID-19 pandemic). During recessions, reductions in working
hours—for companies, these are arbitrated through temporary employment
regulations—become a way to avoid layoffs and make adjustments to the
intensive margin. This flexibility helps, to a large extent, smooth unemployment
rates and has been very effective for addressing temporary or short-term
shocks (Bell and Blanchflower 2013, 2018; Jefferson and Preston 2010; Sum
and Khatiwada 2010). However, chronic underemployment following the Great
Recession, i.e., the emergence of its anticyclical nature, has led analysts to
question whether digitization and the emergence of new forms of employment
have become structural phenomena, maintaining underemployment’s central
role in both research and political action agendas (Valleta et al. 2020;
Borowczyk-Martins and Lalé 2020).
Accordingly, in our analysis, we attempt to provide novel evidence on these
issues in the context of platform-mediated employment.
3. data and methods
3.1. data
Identifying the characteristics of the workforce involved in digital platform
work activities with data from labour force surveys is not straightforward due
to data limitations. On the one hand, there are several difficulties in measuring
nonstandard employment forms by using Labour Force Statistics.5 On the other
hand, the lack of a satellite account for measuring the digital economy is an
added problem (Ahmad & Ribarsky, 2018). This challenge in the measurement
of economic activity has triggered the emergence of new statistical operations
carried out by different institutions using government and big data sources
for taking inventory, identifying practices and monitoring the digital platform
economy, particularly the work mediated by these platforms—see Kilhoffer
(2021), Piasna, (2021), O’Farrell y Montagnier (2020), and Abraham et al.
(2017) for an overview of data sources. In Europe, after a first attempt to provide
an inventory of the population on digital work platforms (Fabo et al., 2017),
the European Commission twice conducted a panel survey (2017 and 2018)
to estimate the prevalence of platform work while obtaining characteristics of
platform workers and their working conditions (Brancati et al., 2018; Pesole
et al., 2019; Brancati et al., 2020). This survey, the COLLaborative Economy
5 Some exceptions are given by the implementation of special modules. The aim of these modules
is to provide users with statistics on a specific topic concerning the labour market by adding a set
of variables to supplement the core Labour Force Survey. In Europe, Denmark, France, Finland, and
Switzerland have carried out operations for measuring the digital platform work by means of ad hoc
modules. For a detailed survey of previous experiences measuring digital platform work, see Kilhoffer
(2021). In this survey, the methods, sources, and the advantages and disadvantages of each approach
are evaluated. The most widely shared view holds that there is no optimal approach to capture all
aspects of digital platform employment. Without challenging this perspective, there are, however,
different methods suitable to measure different facets.
138 María Isabel de Andrés · Emilio Congregado · Concepción Román
and Employment Survey (COLLEEM, hereinafter), has allowed us to analyse
the expansion of the digital labour market in Europe on the basis of recent
empirical work (Pesole et al., 2018; Congregado et al., 2019, 2022; Brancati
et al., 2020).6
3.2. samPle
The European Commission conducted two panel surveys in 2017 and 2018
to estimate the prevalence of digital platform work and characterize digital
platform workers and their working conditions (Brancati et al., 2020). The first
wave, known as COLLEEM I, was completed in 2017 and included information
from 32,409 internet users aged 16 to 74 in 14 European countries, with
19,811 being workers. It collected socio-demographic and labor market
data. The second wave, COLLEEM II, gathered 38,022 responses from 16 EU
Member States, with 26,222 corresponding to workers. Both surveys included
questions for respondents working in digital labor platforms regarding their
characteristics and working conditions. In this work we use only the first wave.7
The first wave of this survey, completed in 2017, contained information
on a total of 32,409 internet users aged between 16 and 74 years old in 14
European countries8, including 19,811 employees and self-employed workers,
while collecting sociodemographic and labour market data. In the survey,
respondents who have been working on digital labour platforms are questioned
about the characteristics and conditions of this platform work.
dePendent variables
Given that our objective is to identify whether there are personal,
sociodemographic or employment characteristics that lead to employment
associated with any of the dimensions of precarity, our first step should
be to review these potential sources of workplace vulnerability. That is,
precarity can be associated with objective reasons, including those that
define underemployment.9 Given the variables included in our sample,
underemployment can therefore be identified as working fewer hours or earning
below-standard pay, with instability or with the perception, subjective in some
6 There is another influential survey for studying platform work in Europe. The survey conducted by
Huws et al. (2017, 2019) is another important source of data available on platform work. However,
this survey seems to overestimate the preponderance of platform workers overall compared to other
surveys (Kilhoffer et al., 2020).
7 Regarding the decision to use only one wave instead of merging the two, it should be noted that
the questionnaire questions varied, preventing the exploration of different dimensions of precarity,
which are the focus of our study.
8 Germany, the Netherlands, Spain, Finland, Slovakia, Hungary, Sweden, the United Kingdom,
Croatia, France, Romania, Lithuania, Italy, and Portugal.
9 To identify underemployment in platform-mediated employment, we use two questions: whether
workers would prefer to work more hours per week in the digital sector if they had more clients
and tasks and whether they feel fairly remunerated for the work they provide through a platform—
underemployment due to a lack of hours vs. underemployment due to a lack of profits.
139
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
cases, of dependency in working conditions or employment protection.10 Since
these different aspects define precarity, different dichotomous variables are
used as dependent variables in the twelve specifications of the estimated
discrete choice models, which should allow us to characterize both the
population groups and the determinants of precarity in platform-mediated
employment.
The precise definitions of the discrete dichotomous variables that we used
to estimate the different unordered binomial models to analyse the probability
of employment are shown in Table A1 of the Appendix.
control variables
The objective of this study is to characterize platform workers by
attempting to identify whether certain types of employment or certain labour
and sociodemographic characteristics make it more likely for employment to
become precarious. This assessment is based on the nature of the work, e.g.,
whether it is face-to-face or not11, and a set of sociodemographic controls, such
as i) the sporadic or principal nature of a job; ii) the degree of work specialization;
iii) the degree of dependence on a given platform; iv) the employment
situation12; v) the initial motivation to access this type of employment; and
vi) previous experience. These variables and some interactions plus a dummy
country – introduced to rule out the existence of idiosyncratic factors – are
included as regressors in the twelve specifications of our model. Depending on
the specification considered, we work with subsamples that move in a range
between 2,192 and 2,418 observations.
10 The COLLEEM includes two items that allow the identification of underemployment by hours
or earnings and a series of items that are associated with precarity in terms of decision-making/
dependency capacity and safety and health.
11 The survey distinguishes between services provided digitally—online—and those that have to be
supplied physically, requiring direct interaction with the service seeker—on location.
12 Regarding the employment situation, the survey covers five different situations: i) employees
who declare a single activity; (ii) self-employed persons who declare themselves to be engaged in a
single activity; (iii) employees who in addition to their main activity carry out a second activity under
some form of self-employment; (iv) “nonemployees”, a heterogeneous group including the inactive
and unemployed; and v) nonwage earners with occasional self-employment including unemployed or
retired persons, students and housewives who have become self-employed.
figure 1. samPle selection by data availability
Notes: PW - Platform worker; sporadic PW – individuals who work on platforms with a lower frequency
than monthly.
140 María Isabel de Andrés · Emilio Congregado · Concepción Román
3.3. methods
Thus, to evaluate our hypotheses, we combine different discrete (binary)
choice models to distinguish the characteristics of groups of platform workers
whose jobs can be associated with some form of precarity.13
For a correct interpretation of the binary model to be estimated, let’s
interpret it in terms of the utility that any pair of exclusive alternatives
provides to an individual, arbitrarily assigning discrete values of 1 and 0 to
them. Under this approach, we assume that the utility associated with each of
the two possible choices depends on a vector of individual and environmental
characteristics, represented by X. Without loss of generality, let’s assume it is a
linear relationship of the form:
In this equation, captures deviations from the average. In this model, an
individual will choose option 1 if the utility of that decision exceeds that of
option 0 and vice versa. Therefore, the endogenous variable in our model can
be defined as,
The baseline model for our estimations can be represented as:
assuming that the distribution function, F, follows a logistic.
4. results
The results of the different estimates are shown in Columns I to XII of Tables
1 and 4. Regarding all columns/specifications, we report the marginal effects of
the different binomial models where the dependent variable is a dichotomous
variable that workers must confront in precarious employment mediated
by a digital platform, defining precarity in several of its forms. In any of the
specifications, the first row includes the predicted probabilities of having a
precarious job, based on the particular aspect used to delimit precarity.
Table A1 shows the specifications that correspond to the two salient
definitions of underemployment. Columns I and II correspond to the estimation
of the probability that a job may be associated with underemployment when
13 The maximum correlation is 0.403 (between the power of decision about what and how to
develop their tasks), and the variance inflation factor values range from 1.03 to 1.625. Hence,
multicollinearity does not seem to be a concern given the size of our sample. The results of the
correlation matrix and the variance inflation factors are available upon request.
141
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
it is defined in terms of hours, while Columns III and IV correspond to the same
estimates when underemployment is defined in terms of earnings.14
Therefore, the first row of Table I indicates that there is a probability over
80% of underemployment in terms of desired work hours in jobs acquired by
a platform, a figure that falls below 9% when we define underemployment in
terms of income.
Underemployment in collaborative employment is significantly characterized
by the form or means used for the provision of a service. Thus, in the first four
specifications (I-IV), there is a greater probability of being underemployed,
regardless of whether it is defined in terms of hours or earnings, for jobs either
fully online or combined with on-site activities. In the latter case, this holds only
when underemployment is defined by insufficient hours, in contrast to those
workers who perform fully on-site activities.
On the other hand, it seems that the intensity of dedication (in the sense of
Pesole et al. 2019) decreases the probability of being underemployed in terms
of earnings. Hence, the probability of developing this form of precarity is lower
for those who have employment in the gig sector as their main activity than for
those who operate in this sector only occasionally or marginally.
The estimates show that less-specialized workers who perform a greater
number of different tasks in contractual employment have a greater probability
of being underemployed as defined by the number of hours, while those
who work for a greater number of platforms also have a greater probability
of underemployment by hours than those who work “exclusively” for a single
platform. Finally, there is no statistically significant evidence that specialization
or exclusivity is linked to a greater probability of underemployment in terms
of earnings.
Among motivations, we find that underemployment by hours is more likely
to occur among workers who have accessed the digital sector due to a lack of
alternatives, a finding that is concentrated in the sample of workers whose job
on a digital platform is their main job or represents a significant percentage of
their work activity (Pesole et al. 2018).15
In experience, it seems that for workers whose first work experience is
platform-mediated, the passage of time decreases their probability of being
underemployed. This effect is especially intense in underemployment by hours,
which is in line with the literature on underemployment that suggests workers
14 Specifications II and IV also incorporate the interactions between educational level and type of
employment as developed through a platform.
15 In Pesole et al. (2018), digital workers are classified according to the number of hours they work
in platform-mediated employment and the income derived from this activity into three categories:
1) platform work is the main activity or a very significant activity, with income at 50% or more of the
total income of the worker and/or the work is done online more than 20 hours a week; 2) platform
work is significant but not a main job, and income from such employment accounts for more than
25% but less than 50% of the individual’s total income and/or work via the platform corresponds
to at least 10 hours per week; and 3) work via platform is not significant, and the income from such
employment does not reach 25% of the worker’s total income.
142 María Isabel de Andrés · Emilio Congregado · Concepción Román
with less work experience are more likely to experience underemployment.
However, this association is usually mediated by the demographic factor of age
(Acosta et al. 2017).
Regarding the effect of some demographic controls, we find no statistically
significant evidence that supports the existence of differences based on
gender or having minor children or the coexistence of the probability of being
underemployed in work that is mediated by a platform. Age, however, has only a
positive effect on the probability of being underemployed, regardless of whether
this is defined by a lack of hours or earnings.
The probability of being underemployed by hours is greater for workers with
intermediate and higher education than for workers with only a basic education.
However, there is no statistically significant evidence of this effect regarding
underemployment in terms of earnings.
On the other hand, the probability of being underemployed in terms of
earnings increases significantly for students and unemployed persons who
occasionally perform a service through a digital platform, although there are
no significant differences concerning such underemployment between self-
employed and salaried workers. This result should not be surprising. While part
of the literature focuses on the idea that in this sector, the notion of false self-
employment or dependent self-employment has spread, these platforms have
become the natural means for finding clients and projects among a host of
freelancers and contractors who actively participate in these platforms’ calls for
tenders (Congregado et al. 2019). Therefore, directly associating self-employment
in this sector with precarity or underemployment is at least questionable.
As we noted above, in Specifications II and IV of Table A1 and in Tables 2 and 3,
we included interactions to capture the potential relationship between the online
or on-site nature of platform-mediated employment and the level of education,
both for underemployment by hours (Table A2) and in terms of earnings (Table
A3). This process entails corroborating or refuting the usual association between
the precarity of on-site platform work—i.e., uberized employment—and a low
level of qualifications. Thus, the separate presentation of the marginal effects
of these interactions should allow us to render a more precise profile of the
effects of education and platform-mediated employment type that allow, where
appropriate, the design of effective actions to address these sources of precarity.
Accordingly, the higher the level of training of on-site workers is, the higher
their probability of being underemployed by hours. Moreover, the group of
workers with basic and secondary education have a higher probability of being
underemployed when they perform on-site activities and when they combine on-
site and online jobs. In other words, the probability of being underemployed by
hours for those workers with a higher education does not markedly differ according
to where they provide their services. However, regarding underemployment in
terms of earnings, the available statistical evidence implies that workers with
secondary and higher education who provide their services online have a greater
probability of being underemployed. This result is in line with recent evidence
(Attewell and Witteveen, 2023).
143
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
Analysing the phenomenon of precarity by considering only a single
manifestation of underemployment would generate a partial and fragmentary
analysis that would define only some of the aspects of vulnerability and
uncertainty that compose the actual, more complex manifestation of precarity
(Olsthoorn 2014). To try to overcome this potential deficiency, in this section, we
present a series of estimates based on alternative definitions of precarity, which
are associated with different aspects linked to bargaining power, protection
systems and working conditions, allowing us to capture some of the dimensions
of the institutional frameworks that are usually associated with precarity. Thus,
by exploiting a battery of questions related to some of these dimensions relevant
to those workers who perform a platform-mediated job, we have been able to
define precarity with new dichotomous variables, accounting for the following:
i) the decision-making capacity of an employee on a platform concerning when
and how many hours to work (Specifications V and VI); ii) the ability of a worker
to choose tasks and how they are carried out (VII and VIII); iii) the safety, hygiene
and health conditions (Specifications IX to XI); and iv) the ability to negotiate
remuneration (XII).
Columns V to XII of Table A416 show the results of the estimates in this
complementary model using 8 different specifications of discrete choice
models to characterize the determinants of the relatively novel aspects of
precarity. Without discussing each of the items separately, there are three
particularly noteworthy results. Firstly, self-employed workers have a more
positive perception of platform employment than salaried workers, as the
former are less likely to experience the causes of precarity in Columns V to
XII. Secondly, and compared to on-site work, it seems that online activities
are associated with less negotiating and decision-making power as well
as less stressful work situations. Thirdly, workers who engage in platform-
mediated employment by necessity, compared to those who engage in
it as an opportunity, do not perceive themselves to be affected by most of
the dimensions associated with precarity (except in relation to stress and
routine tasks). Accordingly, the starting point of the need-based job-seekers,
that is, their lack of other employment opportunities, arguably lowers their
expectations; the employment perceptions among this group of need-based
crowd workers thus seem more favourable than those among workers who
entered platform-mediated employment for other reasons.
5. conclusions
This work addressed the phenomenon of precarity among digital platform
workers by using microdata from a cross-country survey conducted across 14
EU member countries. Different manifestations of precarity were examined to
16 Following the same scheme as Table A1, Table A4 presents the estimates of very similar
specifications. Different dichotomous dependent variables are used according to alternative
definitions based on the aspect of precarity that we consider.
144 María Isabel de Andrés · Emilio Congregado · Concepción Román
provide a definition that is as exhaustive as possible and to identify whether
there are some sociodemographic groups or jobs where the risk or prevalence
of vulnerability is higher.
The novelty of this analysis stems from our treatment of precarity; we
address its different manifestations, including those based on perceptions, to
emphasize the heterogeneity of platform-mediated employment and to verify
whether the direct associations that are usually made between platform work
and precarity are simply mantras based on surveys of expert opinions. We
show that not all forms of precarity show a higher prevalence in the platform
work sector; rather, there are specific population groups and jobs where
precarity has a higher incidence and conditions where precarity can also occur
when employment is mediated by a digital platform.
Furthermore, our results indicate that precarity in this sector might
be fundamentally associated with low-skilled on-site jobs where there is a
relationship of dependence between the applicant and service provider,
including the so-called uberized jobs. However, the mediating role of the
platform in this relationship is not clear. Accordingly, to combat false self-
employment, the correct course of action for political and judicial authorities,
as well as the next European regulations, seems to be identifying dependency
as the key to determining and addressing situations of precarity in the use of
digital platforms for employment.
Our analysis is not exempt from limitations. First, we cannot fully address
some issues with the data to hand since our dataset does not allow us to capture
some dimensions potentially associated with precarity, such as employment
protection, the degree of collective representation, or alternative definitions
of underemployment, such as when skilled workers are in low-income jobs
or underemployment is defined in terms of job satisfaction. More research
is needed to have a complete characterization of precarity in platform work.
Second, our sample does not allow for a deep country-by-country analysis. At
this point, we must be satisfied with contrasting the presence of heterogeneity
by country through dummies. Third, this cross-sectional study would benefit
from the availability of longitudinal data, especially from repeated samples
with the same observed units. Fourth, the availability of richer datasets—
perhaps through special modules of the national labour force surveys—should
allow a characterization of precarity in platforms compared with the rest of
the workers affected by some cause of precarity. This analysis would help
to focus mitigation policies on groups, sectors, and occupations or identify
practices that lead to situations of vulnerability and loss of social protection,
as some scholars suggest (Aloisi 2022). Finally, a micro look at the duration/
survival of individuals in precarious digital platform work and the determinants
of transitions/exit routes to regular/decent jobs would be another natural
extension of this research. Richer (and longitudinal datasets) will not only allow
us to overcome some of the limitations of this work but will also surely fill
some gaps in the future research agenda on precarity in platform-mediated
employment, including the macrodynamics of this phenomenon.
145
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
references
Abraham, K., Haltiwanger, J., Sandusky, K., & Spletzer, J (2017). Measuring
the gig Economy: Current knowledge and Open Issues.Measuring and
Accounting for Innovation in the 21st Century. NBER Working Papers Series,
Working Paper 24950
Acosta, J., Osorno, M.D.P., Rodríguez, O.M. (2018). Underemployment and
Employment among Young Workers and the Business Cycle in Spain: the
Importance of Education Level and Specialization.Journal of Education and
Work, 31(1), 28–46.
Aloisi, A. (2022) Platform Work in Europe: Lessons Learned, Legal Developments
and Challenges Ahead.European Labour Law Journal,13(1): 4-29.
Amuedo-Dorantes, C. (2000). Work Transitions into and out of Involuntary
Temporary Employment in a Segmented Market: Evidence from
Spain.Industrial & Labor Relations Review, 53(2), 309-325.
Arranz, J.M., García-Serrano, C., & Hernanz, V. (2018). Employment
Quality: Are there Differences by Types of Contract?. Social Indicators
Research,137(1), 203-230.
Attewell, P., Witteveen D. (2023). Occupational Marginalization,
Underemployment, and Earnings Inequality among College Graduates.
Research in Social Stratification and Mobility, 83, 00761.
Bell, D.N.F., & Blanchflower D.G. (2013). How to Measure Underemployment?
(No. WP13-7). Peterson Institute for International Economics.
Bell, D.N.F., & Blanchflower D.G. (2018). Underemployment in the US and
Europe(No. w24927). National Bureau of Economic Research.
Bell, D.N.F., & Blanchflower D.G. (2019). The Well-Being of the Overemployed
and the Underemployed and the Rise in Depression in the UK.Journal of
Economic Behavior & Organization,161, 180-196.
Berg, J., & Rani, U. (2018).Digital Labour Platforms and the Future of Work:
Towards Decent Work in the Online World. Geneva: International Labour
Organization.
Borowczyk-Martins, D., & Lalé, E. (2020). The Ins and Outs of Involuntary Part-
Time Employment. Labour Economics, 67: 101940.
Bogliacino F., Codagnone C., Cirillo V., & Guarascio, D. (2020). Quantity and
Quality of Work in the Platform Economy. In Handbook of Labor, Human
Resources and Population Economics, edited by K. Zimmermann. Champ.:
Springer.
Büchteman, C.F., & Quack, S. (1990). How Precarious is ‘Non-Standard’
Employment? Evidence for West Germany. Cambridge Journal of
Economics, 14(3): 315-329
Brancati, M.C., Pesole, A., & Fernandez Macias, E. (2019).Digital Labour
Platforms in Europe: Numbers, profiles, and employment status of platform
workers(No. JRC117330). Joint Research Centre.
146 María Isabel de Andrés · Emilio Congregado · Concepción Román
Brancati, U., Pesole, A., & Fernández-Macías, E. (2020). New Evidence on
Platform Workers in Europe.Results from the Second COLLEEM survey.
Luxembourg: Publications Office of the European Union.
Congregado, E., de Andrés, M.I., & Roman, C. (2019). Explorando la
heterogeneidad en el empleo mediado por plataformas digitales: Intensidad
y localización.Revista de Economía Laboral,16(2): 69-103.
Congregado, E., de Andrés, M.I., Nolan, E, & Roman, C, (2022). Heterogeneity
Among Self-employed Digital Platform Workers. Evidence from Europe.
International Review of Entrepreneurship, 20(1): 45-67.
De Groen, W.P., Kilhoffer, Z., Lenaerts, K., & Salez, N (2017). The Impact of the
Platform Economy on Job Creation.Intereconomics,52(6): 345-351.
De Stefano, V. (2016). The Rise of the Just-in Time Workforce: On Demand
Work, Crowdwork, and Labor Protection in the Gig Economy. Comparative
Labor Law and Policy Journal, 37(3): 461-471.
De Stefano, V., & Aloisi, A. (2019). Fundamental Labour Rights, Platform
Work and Human Rights Protection of Non-Standard Workers. InResearch
Handbook on Labour, Business and Human Rights Law. Edward Elgar
Publishing.
Drahokoupil, J., & Piasna, A. (2017). Work in the Platform Economy: Beyond
Lower Transaction Costs. Intereconomics 52: 335–340. https://doi.
org/10.1007/s10272-017-0700-9
Fabo, B., Karanovic, J., & Dukova, K. (2017). In Search of an Adequate
European Policy Response to the Platform Economy.Transfer: European
Review of Labour and Research,23(2): 163-175.
Feldman, D.C. (1996). The Nature, Antecedents and Consequences of
Underemployment.Journal of Management, 22: 385–407.
Findeis, J.L., Jensen, L., & Wang, Q. (2000). Underemployment Prevalence and
Transitions in the US Nonmetropolitan South.Southern Rural Sociology,16:
125-147.
Congregado, E., Garcia-Clemente, J., Rubino, N., & Vilchez, I. (2024). Testing
Hysteresis for the US and UK Involuntary Part-Time Employment.Applied
Economics, 1-21.
García-Pérez, C., Prieto-Alaiz, M., & Simón, H. (2020). Multidimensional
Measurement of Precarious Employment Using Hedonic Weights: Evidence
from Spain. Journal of Business Research, 113: 348-359,
Graham, M., Hjorth, I., & Lehdonvirta, V. (2017). Digital Labour and
Development: Impacts of Global Digital Labour Platforms and the Gig
Economy on Worker Livelihoods.Transfer: European Review of Labour and
Research,23(2), 135–162.
Hussmanns, R., Mehran, F., & Verma, V. (1990). Surveys of Economically
Active Population, Employment, Unemployment, and Underemployment:
an ILO Manual on Concepts and Methods. Geneve: International Labour
Organization.
Huws, U., Spencer, N.H., Syrdal, D.S., Holts, K. (2017). Work in the European
gig Economy: Research Results from the UK, Sweden, Germany, Austria,
147
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
the Netherlands, Switzerland and Italy. FEPS – Foundation for European
Progressive Studies.
Huws, U., Spencer, N.H., Coates, M., & Holts, K. (2019). The Platformisation
Of Work In Europe: Highlights from Research in 13 European Countries.
Foundation for European Progressive Studies, July.
Kalleberg, A.L. (2009). Precarious Work, Insecure Workers: Employment
Relations in Transition.American Sociological Review,74(1): 1-22.
Kässi, O., & Lehdonvirta, V. (2018). Online Labour Index: Measuring the Online
Gig Economy for Policy and Research.Technological Forecasting and Social
Change,137: 241-248.
Kauhanen, M., & Nätti, J. (2015). Involuntary Temporary and Part-Time Work,
Job Quality and Well-Being at Work.Social Indicators Research, 120(3):
783-799.
Kenney, M., & Zysman, J. (2016). The Rise of the Platform Economy.Issues in
Science and Technology,32(3): 61-69.
Kilhoffer, Z., De Groen, W.P., Lenaerts, K., Smits, I., Hauben, H., Waeyaert,
W., Giacumacatos, E., Lhernould, J.P., & Robin-Olivier, S. (2020). Study
to Gather Evidence on the Working Conditions of Platform Workers,
Directorate-General for Employment, Social Affairs and Inclusion.
Kilhoffer, Z. (2021). State-of-the art. Data on the Platform Economy, Deliverable
n°12.3, Leuven, InGRID-2 project 730998 – H2020.
Koutsimpogiorgos, N., Van Slageren, J., Herrmann, A.M., Frenken, K. (2020).
Conceptualizing the Gig Economy and Its Regulatory Problems.Policy &
Internet,12(4): 525-545.
Kretsos, L., & Livanos, I. (2016). The Extent and Determinants of Precarious
Employment in Europe.International Journal of Manpower, 37(1): 25-43.
Kuhn, K.M., & Maleki, A. (2017). Micro-entrepreneurs, Dependent Contractors,
and Instaserfs: Understanding Online Labor Platform Workforces.Academy
of Management Perspectives, 31(3):183-200.
Larsson, A., & Teigland, R. (2020).The Digital Transformation of Labor. London:
Taylor & Francis.
Lehdonvirta, V. (2018). Flexibility in the Gig Economy: Managing Time on Three
Online Piecework Platforms. New Technology, Work and Employment,
33(1): 13–29.
Lepanjuuri, K., Wishart, R., & Cornick, P. (2018). The Characteristics of those
in the Gig Economy.UK Department for Business, Energy and Industrial
Strategy.
Malo, M.A. (2018). Nuevas formas de empleo: del empleo atípico a las
plataformas digitales.Papeles de Economía Española, 156:146-158.
Maynard, D.C., & Feldman, D.C. (Eds.) (2011).Underemployment: Psychological,
Economic, and Social Challenges. Springer Science & Business Media.
McKee-Ryan, F.M., & Harvey, J. (2011). I Have a Job, but...: A Review of
Underemployment.Journal of Management,37(4):962-996.
148 María Isabel de Andrés · Emilio Congregado · Concepción Román
Monastiriotis, V., & Laliotis, I. (2020). Labour Market Vulnerability and Labour
Market Outcomes during the Economic Upswing, KE 0419795 EN,
Publications Office of the European Union, Luxembourg, 2020.
Montgomery, T., & Baglioni, S. (2020). Defining the Gig Economy: Platform
Capitalism and the Reinvention of Precarious work.International Journal of
Sociology and Social Policy 48(4): 597-600.
O’Farrell, R., Montagnier, P. (2020). Measuring Digital Platform-Mediated
Workers.New Technology, Work and Employment,35(1): 130-144.
Olsthoorn, M. (2014). Measuring Precarious Employment: A Proposal for Two
Indicators of Precarious Employment Based on Set-Theory and Tested with
Dutch Labor Market-Data. Social Indicators Research, 119(1): 421-441.
Pantea, S. (2021). Self-Employment In The EU: Quality Work, Precarious Work
Or Both?.Small Business Economics, 57(1): 243-263.
Pesole, A., Brancati, U., Fernández-Macías, E., Biagi, F., & Gonzalez-Vazquez,
I. (2018) Platform Workers in Europe.Luxembourg: Publications Office of
the European Union.
Pesole, A., Fernandez-Macias, E., Brancati, C.U., & Herrera, E.G. (2019).
How to Quantify what is not Seen? Two Proposals for Measuring Platform
Work(No. 2019/01). JRC Working Papers Series on Labour, Education and
Technology.
Piasna, A. (2021). Measuring the Platform Economy: Different Approaches to
Estimating the Size of the Online Platform Workforce.A Modern Guide To
Labour and the Platform Economy, 66-80.
Rodgers, G., & Rodgers, J. (1989). Precarious Jobs in Labour Market
Regulation. The Growth of Atypical Employment in Western Europe,
Genève: International Labor Organization.
Román, C., Congregado, E., & Millán, J.M. (2011). Dependent Self-Employment
as a Way to Evade Employment Protection Legislation.Small Business
Economics,37(3): 363-392.
Schor, J.B., Attwood-Charles, W., Cansoy, M., Ladegaard, I., & Wengronowitz,
R. (2020). Dependence and Precarity in the Platform Economy.Theory and
Society,49(5): 833-861.
Stanford, J. (2017). The Resurgence of Gig Work: Historical and Theoretical
Perspectives.The Economic and Labour Relations Review,28(3): 382-401.
Stewart, A.., & Stanford, J. (2017). Regulating Work in the Gig Economy: What
Are the Options?The Economic and Labour Relations Review,28(3): 420-
437.
Tam, H. (2010). Characteristics of the Underemployed and the Overemployed
in the UK.Economic & Labour Market Review,4(7): 8-20.
Todolí-Signes, A. (2017). The ‘Gig Economy’: Employee, Self-Employed or the
Need for a Special Employment Regulation?Transfer: European Review of
Labour and Research,23(2): 193-205.
Valletta, R.G., Bengali, L., & Van der List, C. (2020). Cyclical and Market
Determinants of Involuntary Part-Time Employment.Journal of Labor
Economics,38(1): 67-93.
149
Precariousness in emPloyment mediated by digital Platforms. evidence from euroPe
revista de economía mundial 66, 2024, 129-149
Vosko, L. (2006). Precarious Employment: Understanding Labour Market
Insecurity in Canada. Montreal: McGill-Queen’s University Press.
Vosko, L,.Macdonald, M.,Cambell, I. (Eds) (2010). Gender and the Countours
of Precarious Employment.London and New York, NY: Routledge.
Webster, J. (2016). Microworkers of the Gig Economy: Separate and Precarious.
New Labor Forum25(3): 56-64.
Weststar, J. (2009). Worker Control as a Facilitator in the Match Between
Education and Jobs.British Journal of Industrial Relations,47(4): 723-740.
Weststar, J. (2011). A Review of Women’s Experiences of Three Dimensions of
Underemployment. In Underemployment, edited by Maynard D. and D.
Feldman, 105-125. Springer, New York, NY.
anexo
Las tablas A1, A2, A3, A4, A5 y A6 se encuentran en formato .pdf y .html
en la siguiente dirección:
https://www.uhu.es/publicaciones/ojs/index.php/REM/article/view/8065/7013
