This is a preprint version of the THE THIRD-LEVEL DIGITAL DIVIDE: WHO
BENEFITS MOST FROM BEING ONLINE? Published in Communication and
Information Technologies Annual: Digital Distinctions and Inequalities. Studies in
Media and Communications, Volume 10, 29-53. See for published version:
http://www.emeraldinsight.com/doi/abs/10.1108/S2050-206020150000010002
3
THE THIRD-LEVEL DIGITAL
5
DIVIDE: WHO BENEFITS MOST
7
FROM BEING ONLINE?
9
11
Alexander J. A. M. van Deursen and Ellen J. Helsper
13
15
ABSTRACT
17
Purpose
-
Research into the explanations of digital inclusion has moved
from investigations of skills and usage to tangible outcomes, what we label
19
here as the third-level digital divide. There is a lack of theoretical develop-
ment about which types of people are most likely to benefit. Understanding
21
how achieving outcomes of internet use is linked to other types of (dis)
advantage is one of the most complex aspects of digital inclusion research
23
because very few reliable and valid measures have been developed. In the
current study we took a first step toward creating an operational frame-
25
work for measuring tangible outcomes of internet use and linking these to
the inequalities identified by digital divide research.
27
Methodology/approach
-
After having proposed a classification for
internet outcomes, we assessed these outcomes in a representative sample
29
of the Dutch population.
31
Findings
-
Our overall conclusion in relation to the more general rela-
tionship between offline resources and third-level digital divides is that
33
35
Communication and Information Technologies Annual: Digital Distinctions and Inequalities
Studies
in
Media
and
Communications,
Volume
10,
29
-
53
Copyright 2015 by Emerald Group Publishing Limited
All rights of reproduction in any form reserved
39
ISSN: 2050-2060/ doi:10.1108/S2050-206020150000010002
29
37
30
A. J. A. M. VAN DEURSEN AND E. J. HELSPER
1
the internet remains more beneficial for those with higher social status,
not in terms of how extensively they use the technology but in what they
3
achieve as a result of this use for several important domains.
5
Implications
-
When information and services are offered online, the
number of potential outcomes the internet has to offer increases. If indi-
7
viduals with higher social status are taking greater offline advantage
from
digital engagement than their lower status counterparts, existing
9
offline inequalities could potentially be acerbated.
Keywords: Outcomes; third-level digital divide; internet use; digital
11
inequality; social inequality
13
15
INTRODUCTION
17
By now a vast array of studies have illuminated the consequences of digital
19
inequalities for many different offline activities and life realms. Examples
include research on political participation, educational attainment, and
21
employment outcomes. What the field lacks is a comprehensive and sys-
tematic study which charts gaps in offline outcomes among sociodemo-
23
graphic and socioeconomic groups across multiple domains of activity.
More specifically, we know little about such gaps in societies where internet
25
access is very widely diffused within the population.
Such a task is important if we want to gain a deeper and broader under-
27
standing of the third-level digital divide and its repercussions for offline
inequalities. The third-level digital divide concerns disparities in the returns
29
from internet use within populations of users who exhibit broadly similar
usage profiles and enjoy relatively autonomous and unfettered access to ICTs
31
and the internet infrastructure. Third-level divides, therefore, relate to gaps in
individuals’ capacity to translate their internet access and use into favorable
33
offline outcomes. Research into the third-level divide, therefore, seeks to
determine who benefits in which ways from internet use in terms of a broad
35
range of offline outcomes (Amichai-Hamburger, McKenna, & Tal, 2008;
Stern, Adams, & Elsasser, 2009; van Deursen, van Dijk, & Helsper, 2014).
37
Research into the third-level divide has taken many steps forward in
recent years, but it has not yet attempted to chart gaps in returns from inter-
39
net usage across multiple life realms within a uniformly wired society where
internet access is almost universal. Advancing this research necessitates
The Third-Level Digital Divide
31
1
linking types of digital engagements to specific offline life realms such as eco-
nomic, social, and political life realms. Quantitative research into the third-
3
level divide stands to gain, if specific digital engagements can be linked to
outcomes in particular life realms, a deeper understanding of the mechanisms
5
translating internet use into specific offline outcomes (e.g., Stern et al., 2009).
Such an exercise would also afford the opportunity for the development of
7
theoretically informed classificatory schemes by which researchers can sort
internet users in terms of the likely offline benefits accruing to specific types
9
of internet use. Rather than assuming that more digitally advantaged users
will automatically enjoy greater offline benefits across all life realms, the
11
strength and character of the links between skills, online activities, and
offline outcomes should be treated as factors which can potentially vary
13
across domains and fields of activity. Indeed, where existing digital divide
research does touch on the third-level divide, it suggests that, as a rule, inter-
15
net use and online activities will confer greater benefits to internet users
in life realms where the user already has significant resources at his or her
17
command
(DiMaggio & Hargittai, 2001; Hargittai & Hinnant, 2008;
Helsper, 2012; van Deursen & van Dijk, 2011, 2014; van Dijk, 2005).
19
The study presented in this paper should be considered as a preliminary
step toward devising an operational framework useful for charting the con-
21
tours of the third-level digital divide in a society where internet access is
near-universal. It also will serve to elucidate some of the mechanisms
23
through which internet usage is converted into offline benefits. It does so
by identifying which groups derive greater and lesser offline returns, given
25
particular levels of internet usage, across distinct economic, political, and
institutional life realms. We therefore ask: What are the returns on internet
27
use for particular sociodemographic groups identified by digital divide
research
and how are these returns linked to particular usage patterns?
29
We hypothesize at the outset that greater returns will accrue to those more
favorably situated users.
31
33
THEORETICAL BACKGROUND
35
Digital Divides
37
Conceptualized at the dawn of the digital age, the notion of the first-level
39
digital divide trains attention on individuals’ access to the ICT infrastruc-
ture, including such dimensions as autonomy and continuity of access
32
A. J. A. M. VAN DEURSEN AND E. J. HELSPER
1
(Newhagen & Bucy, 2005; van Dijk, 2005). As more and more people
obtained access to this infrastructure, second-level divides in skills and
3
usage patterns became more evident and drew more attention from
researchers (e.g., Dimaggio et al., 2004; Katz & Rice, 2002; Selwyn, 2006;
5
van Dijk, 2005; Witte & Mannon, 2010; Zillien & Hargittai, 2009). Studies
of second-level digital divides have now provided, for example, useful clas-
7
sifications in terms of the types of skills needed to use ICTs and the types
of activities people perform online (e.g., Blank & Groselj, 2014; Kalmus,
9
Realo, & Siibak, 2011; van Deursen & van Dijk, 2011, 2014; Warschauer,
2003) and how these digital divide aspects interact (e.g., Livingstone &
11
Helsper, 2007; van Deursen & van Dijk, 2015). It is well known that first-
level and second-level digital divides have important implications for offline
13
outcomes in societies or groups where access is unevenly distributed
(Robinson, 2009; Witte & Mannon, 2010).
1
15
The third-level digital divide differs from first-level and second-level
divides, inasmuch as the first-level digital divide concerns differences in
17
infrastructural access, and second-level digital divides have to do with dif-
ference in skills and usage patterns (Hargittai, 2002). Insufficient skills
19
have been found to play a role in limiting success or efficiency in the
undertaking of specific online tasks. In societies such as the Netherlands
21
with near-universal internet access, however, third-level digital divides
have
become increasingly salient. Research into third-level divides pre-
23
sume that, even among users with autonomous and unlimited access to
the
ICT infrastructure, there will be important differences in their profi-
25
ciency in enlisting digital resources for the achievement of specific objec-
tives. Even when two users have high-quality autonomous access and
27
adequate skills, they may not obtain the same returns on their internet use
(Stern et al., 2009; van Deursen et al., 2014). Moreover, individuals who
29
consistently convert their internet use into high offline returns such as
earnings may benefit from a feedback effect where greater economic
31
resources enable them to further develop their internet skills. For exam-
ple, someone gaining a better job through the use of the internet might
33
have access to an increased wage which in turn can be used to get better
access, improve their skills and, thus, buy products cheaper online.
35
The outcomes achieved from internet use provide feedback into someone’s
offline status which then again influences the digital inclusion factors as
37
illustrated in Fig. 1.
In this paper, we focus on measuring the benefits that result from inter-
39
net use across multiple life realms and how these benefits relate to member-
ship in specific sociodemographic groups. These outcomes are rarely
The Third-Level Digital Divide
33
1
3
5
7
9
11
Fig. 1. A Model for Replications of Inequalities in a Digital Society.
13
Source: Adapted from Helsper (2012) and van Dijk (2005).
15
measured in large-scale, population-wide surveys that aim to provide a
17
broader understanding of what people gain from internet use. Some studies
focused on the so-called “opportunity divide” (e.g., Akca, Sayili, &
19
Esengun, 2007; Mossberger, Tolbert, & Stansbury, 2003; Stern et al., 2009).
These studies are mostly concerned with tracking different types of internet
21
use, rather than identifying the offline outcomes that result from these
online
usage patterns. Many other studies focus only on one particular
23
type of outcome, for example on establishing social networks (e.g., Boase,
Horrigan, Wellman, & Rainie, 2006) or increasing political participation
25
(e.g., Sylvester & McGlynn, 2010). As a result, the actual implications of
internet use in terms of real opportunities in everyday life are increasingly
27
important to digital divide research.
When studies measure outcomes they either focus on a very narrow
29
range of indicators of one particular type (e.g., Boase et al., 2006;
Sylvester
& McGlynn, 2010) or collapse a whole range of outcomes
31
together without specifying why different types of outcome items were
included (Stern et al., 2009). To enrich quantitative work in this area, we
33
gauge offline outcomes across multiple fields of activity as effects of speci-
fic types of internet usage, laying the groundwork for generalizations
35
about the linkages between pathways from attitudes, access, skills, use,
and
digital engagements to offline outcomes. However, before we begin
37
this exercise, we need to settle on an analytical framework which allows us
to categorize the relevant fields of offline activity. In this regard, systematic
39
theorization-based conventional understandings of offline inequality can
serve as a useful starting point.
Digital
inclusion
indicators
Skills
Offline
resources
Access
Use
Offline
outcomes
of online
activities
Moti-
vation
Socio-
demo-
graphics
