You are on page 1of 23

Journal of Technology in Human Services

ISSN: 1522-8835 (Print) 1522-8991 (Online) Journal homepage: http://www.tandfonline.com/loi/wths20

Perpetuating Old Exclusions and Producing New


Ones: Digital Exclusion in an Information Society

Yu Cheung Wong , Chi Kwong Law , John Yat Chu Fung & Jolie Chi Yee Lam

To cite this article: Yu Cheung Wong , Chi Kwong Law , John Yat Chu Fung & Jolie Chi
Yee Lam (2009) Perpetuating Old Exclusions and Producing New Ones: Digital Exclusion
in an Information Society, Journal of Technology in Human Services, 27:1, 57-78, DOI:
10.1080/15228830802459135

To link to this article: http://dx.doi.org/10.1080/15228830802459135

Published online: 28 Jan 2009.

Submit your article to this journal

Article views: 162

View related articles

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=wths20

Download by: [Liverpool John Moores University] Date: 24 October 2016, At: 01:19
Journal of Technology in Human Services, 27:5778, 2009
Copyright # Taylor & Francis Group, LLC
ISSN: 1522-8835 print=1522-8991 online
DOI: 10.1080/15228830802459135

Perpetuating Old Exclusions


and Producing New Ones: Digital
Exclusion in an Information Society

YU CHEUNG WONG and CHI KWONG LAW


The University of Hong Kong, Pokfulam, Hong Kong

JOHN YAT CHU FUNG


Director, Information Technology Resource Centre, Hong Kong Council of Social Service,
Wanchai, Hong Kong

JOLIE CHI YEE LAM


Senior Research Assistant, The University of Hong Kong, Pokfulam, Hong Kong

This article presents a study that measures the degree of digital


exclusionor, conversely, the degree of digital inclusionin Hong
Kong, a developed city in East Asia. Governments in the region are
among the most active in the developed world in pushing ahead in
developing knowledge economies and information societies. The
major concern is to improve=maintain their competitiveness in
the new knowledge economy created by the process of globalization
and the advancement in information technology. Many countries
in the region have established themselves in the top ranks of a
number of indexes and measurements comparing digital readi-
ness, digital access, information and communication technology
penetration, and others. However, not all the citizens in the region
share the benefits and promises of the information society. People
who are traditionally disadvantaged, such as the elderly and those
on a low income, are further excluded from the information
society. Such exclusion affects other social groups as well. This
study creates a new digital inclusion index to measures the degree
of inclusion of various disadvantaged groups in an information
society. Data regarding seven disadvantaged groups were collected
through a household survey (N 2,312). The index captures

Address correspondence to Yu Cheung Wong, PhD, Department of Social Work and


Social Administration, The University of Hong Kong, Pokfulam, Hong Kong, SAR. E-mail:
ssycwong@hku.hk

57
58 Y. C. Wong et al.

information about access, knowledge, usage, and affordability


in information and communication technology of the disadvan-
taged in comparison with mainstream society (N 284).

KEYWORDS digital divide, digital inclusion index, disadvantaged


groups, ICT

In the previous decade, we have experienced one of the greatest transforma-


tions in history in the way societies create, store, distribute, and use informa-
tion. Many of us are now living in an information society, enjoyingand
at times complainingabout the changes it has brought in working, learning,
interacting with others, and entertaining ourselves. The information society
has greatly benefited many people. However, there are still many people
who cannot benefit from what electronic information promises because they
do not have access to or do not know how to use the technology and tools.
This is usually referred to as the digital divide, and in many ways it is a
manifestation or exacerbation of existing divisions due to income, age, edu-
cation level, etc. Many countries consider it important to bridge the divide
through promoting public education and training programs. For such efforts
to be effective it is vital to have an instrument to measure and keep track of
the size of the divide. This will enable the identification of the most excluded
groups and enable the development of more targeted programs of which
overall effectiveness can be evaluated. The Digital Inclusion Index described
in this paper attempts to provide such a measurement.

LITERATURE REVIEW

The digital divide is about the difference in rates of access to computers and
the Internet (U.S. Department of Commerce, 2000, p. 2), or is commonly
understood as the gap between information and communications technology
(ICT) haves and have-nots (Sciadas, 2002, p. 2). Despite the lack of any
serious academic definition, the concept has been widely adopted by
governments (World Summit on the Information Society, 2003; European
Commission, 2005). Such differences result in the exclusion of those who
do not have access or have only limited access preventing them from taking
part in the information society. This new form of exclusion not only limits
individuals from sharing the benefits of the information society and maximiz-
ing their life chances but also hampers the overall development of a knowl-
edge economy, which is vital for the future of societies in a highly globalized
and competitive world.
Internet connectivity has grown rapidly since it began to be more
widely used by the general public (from around 1995). In 2007, there were
1.7 billion Internet users, who have access to the Internet and have basic
Digital Exclusion in an Information Society 59

knowledge to use it, all over the world, representing only 17.8% of the
world population (Internet World Stats, 2007b). However, Europe and
North America together, although their population constitutes only 17.4%
of the total world population, accounted for nearly half (47.2%) of all
Internet users on the planet.
In Hong Kong, the number of computer users aged 10 and older has
increased from 43.1% in the first quarter of 2000 to 62.9% in 2006. Most com-
puter users also used the Internet (Census and Statistics Department, 2007).
In China, the number of Internet users was 162 million in July 2007, repre-
senting an almost tenfold growth since July 2000, when they numbered only
16.9 million (China Internet Network Information Centre, 2000, 2007). The
diffusion rate of Internet technology is faster than many other technologies
such as fixed line telephones, television, and even mobile phones (U.S.
Department of Commerce, 2004, p. 6).
Since the knowledge economy is vital for future success, societies are
competing vigorously in their ability to develop and attract human capital
through investment in education, research, and development, accelerating
the globalized information society (Organisation for Economic Co-operation
and Development, 2003, p. 7). Once developed societies started to compete
in building knowledge economies and information societies, a widening gap
emerged between rich and poor countries in the use of ICT, the very back-
bone that makes an information society possible. Take, for example, indivi-
duals in a rich country who can have more international data flow into their
own computers through the Internet than the whole population of a poor
country: A single 100 Mbit=s broadband user in a leading broadband coun-
try such as Japan has access to as much international connectivity as the 45
countries with the lowest international connectivity combined (Organisation
for Economic Co-operation and Development, 2004, p. 299).
Furthermore, the quality of Internet access differs greatly. For the top 20
countries, there were altogether 80.2 million DSL broadband subscribers
while the rest of the world had only about 5.1 million (Internet World Stats,
2007a). Even if Internet access is available, it could be prohibitively expen-
sive in some countries, especially in those whose telecommunication sector
is dominated by a single provider. A recent report by the Organization for
Security and Co-operation in Europe indicated that an unlimited dial-up plan
costs about 482 ($111) a month in Kazakhstan, where the average monthly
wage is 4292 ($399), and an unlimited 1.5Mbps connection costs up to
42,458 ($3,355) a month (Moller & Amouroux, 2007).
The disparities in access to telecommunication technology (including
mobile phone, computer, and Internet) across countries have become an
international concern especially for poor countries. Limited access to these
new technologies because of poor infrastructure, low levels of computer
literacy (as well as general literacy), and lack of affordability further prohibit
poorer countries from benefiting from the knowledge economy and
60 Y. C. Wong et al.

globalization. Indicators have been developed to gauge the level of dispari-


ties among countries. For example, the International Telecommunication
Union (ITU), a specialized agency of the United Nations, has developed
the Digital Access Index in 2003 (International Telecommunication Union,
2003) to measure such disparities. In this index, Hong Kong ranks seventh
in the table, second only to South Korea in the Asian Pacific region, which
ranks fourth in the world (International Telecommunication Union, 2003).
The disparities in access within countries have become a major concern
of the developed countries because of the effect on individual well-being as
well as the countries survival in the new knowledge economy. In 2000, the
European Council considered that the emergence of new information and
communication technologies constitutes an exceptional opportunity,
provided that the risk of creating an ever-widening gap between those
who have access to the new knowledge and those who do not is avoided
(Employment and Social Policy Council, 2000). In 2005, after years of effort
to reduce the disparities within their countries, they still worried about the
very slow development [in access] concerning low income and less
educated groups. . .becoming a history of ever evolving delays and=or
permanent exclusion. . . (European Commission, 2005, p. 12).
In 1995, 1998, and 2000, the United States Department of Commerce
released its Falling through the Net reports in which unequal access asso-
ciated with rural residence, race, education level, gender, and age was exam-
ined. In 2000, the Organisation for Economic Co-operation and Development
(OECD) released Understanding the Digital Divide (Organisation for
Economic Co-operation and Development, 2001), which examined the
unequal distribution of access throughout OECD countries (Organisation
for Economic Co-operation and Development, 2004). Initially, affordability
and literacy (which affects confidence to use the new technology) are major
barriers to access within developed countries. However, with the falling price
of personal computers and Internet services and better infrastructure due to
market competition in recent years, affordability and access to facilities (free
or low cost access points have been set up in the neighborhood) have
become a less important factor. The main barriers seem related to literacy,
language, lack of cultural diversity, lack of local content (European Commis-
sion, 2005, p. 15). In addition, as high quality but more expensive broadband
access has become popular, broadband divide (Organisation for Economic
Co-operation and Development, 2004, p. 298) has begun to emerge. This
divide consists of a disparity in the quality of information services, both in
terms of the transmission speed and the content of the information. This will
ultimately affect the usage of electronic information and thus the level of
inclusion in the information society, especially for disadvantaged groups.
In Hong Kong, despite a high level of computer and Internet usage,
usage is still very low among the elderly and less educated. In particular,
older persons have a computer usage rate of 25.9% for those aged 5564
Digital Exclusion in an Information Society 61

and 5.3% for those aged 65 or above, whereas the overall average of all
persons aged 10 and above is 52.9% (Census and Statistics Department,
2007, p. 24). Women also have a slightly lower usage rate than men
(60.1% versus 65.8%) (Census and Statistics Department, 2007, p. 26).
Besides, lower income households are less likely to have a computer at
home (Census and Statistics Department, 2007, p. 23). The percentage of
computer users is the highest for persons with postsecondary education
or above, at 95.3%. The corresponding rates for those with secondar-
y=matriculation educational attainment and primary educational attainment
and lower were 72.4% and 20.8% respectively (Census and Statistics
Department, 2007, p. 28).

CONCEPTUAL FRAMEWORK FOR UNDERSTANDING


DIGITAL INCLUSION

Digital inclusion refers to the extent to which the disadvantaged segments of


a society can have equal opportunity to take part in the information society.
Promoting digital inclusion aims to close such opportunity gaps. The prere-
quisite for having an opportunity to take part in the information society is
having access to information and communication technology. The degree
to which disadvantaged people actually use these tools to engage in the
information society also reflects the extent to which equal opportunity exists.
Conceptually, accessibility and actual usage are interwoven and serve as
important indicators of whether disadvantaged groups have equal opportu-
nity to take part in the information society.
There are several important factors that affect the level of accessibility
and usage of ICT. Affordability and ability to use ICT are obvious candidates,
without which an individual has little chance of enjoying the benefits of the
information society. Free or low cost public access could compensate for the
affordability factors, and having useful and important content for certain dis-
advantaged groups also helps to increase the incentives to learn and use ICT
(Wong, Law, Fung, & Lam, 2006).
Figure 1 presents the major factors that are important to accessibility and
the degree of usage. Such factors include affordability, ICT skills=literacy,
availability of the essential technology (for example, in some neighborhoods,
no Internet connection is available), the existence of the desired applications
and content and other sociocultural factors (for example, being unwilling to
use imported technology).
Some of the factors can be easily measured and in some cases obtained
from secondary sources of data. For example, affordability, which is mainly
represented by the household income of its members, can be readily
ascertained using government statistics. Other factors, such as the existence
of desired content and applications and sociocultural factors are subjective
62 Y. C. Wong et al.

FIGURE 1 Conceptual framework of digital inclusion.

perceptions and difficult to capture. In this study, we focused on the afford-


ability and capacity (knowledge and skills) to use ICT and ignored availabil-
ity since, by and large, infrastructure and essential devices were readily
available in the local market. In order to measure the two main dimensions
of digital inclusion and related factors they were operationalized into
relevant indicators.

Accessibility
Accessibility refers to the accessibility of personal computers and Internet
connections, at home and in other locations so that individuals can connect
to the Internet to use its services. Increasingly, the quality of technology, for
example, having a broadband connection, and higher processing power and
storage capacity affects the intention and satisfaction of using technology-
enabled services. Thus this will be included as one of the indicators of acces-
sibility. In a relatively developed city, having a home computer is almost
guaranteed for those in the mainstream. In addition, since public access is
generally available in the local communities, it therefore makes more sense
to study the disparity in the opportunity to engage in the information society
by focusing on the accessibility to home computers and Internet connections
instead of to public or other non-home computers. Thus the comparison will
focus on the extent to which individuals (a) have a home computer (desktop
and=or notebook computer), (b) have home computers connected to the
Internet, (c) have a good quality computer, (d) have a broadband connec-
tion, and (e) have a computer at home that they can use without permission.

Usage
Usage refers to the actual usage of the computer and services enabled by
information and communication technology. Such services include those
Digital Exclusion in an Information Society 63

that are commonly available and are important to the functioning of indi-
viduals, including searching for information through the Internet, commu-
nicating with others (unsynchronized, such as sending and receiving
e-mail, short messages, bulletin board messages; and synchronized, such
as interacting with others in chat rooms, using instant messages), using
e-government services (e.g., venue booking, license application), con-
ducting online business transactions (e.g., auctions, online payments,
online shopping), and producing websites to share information with
others. It is also important to capture the amount of time spent in using
those technology-enabled services. The indicators include (a) having ever
used a computer, (b) history (measured in the number of years) of having
used a computer, (c) frequency of using the Internet, and (d)
average duration (number of hours) of Internet usage each time (in the
past three months).

Computer Skills=Literacy
Computer skills=literacy or knowledge refers to individuals self-assessment
of their knowledge about the uses of a computer and the Internet, their
ability (self-assessed) to use the technology-enabled services mentioned
previously, and the number of hours of training received in using compu-
ter=Internet. Computer skills=literacy or knowledge tend to be highly
correlated with general literacy and educational attainment (Sciadas, 2002;
U.S. Department of Commerce, 2000). In order to make international com-
parisons, educational attainment, information about which is readily avail-
able from all countries, is taken as a reliable and accurate proxy for
computer skills=literacy. In this study, the indicators for computer skills=lite-
literacy include (a) overall knowledge (self-perception) in using compu-
ter=Internet, (b) ability (self-assessment) to engage in various kinds of
online activities, and (c) amount of training (number of hours) received
in using a computer=Internet.

Affordability
Affordability refers to the ability to pay for the devices and services in order
to engage in the information society through using various technology-
enabled services. Since by and large the market is open to all (except in
some localities, where Internet service providers [ISPs] are not free to
operate), the household income reflects the affordability of devices and
services. Household income is preferred to individual income because
individuals with little or no income (e.g., older adults, homemakers)
could have good affordability if they come from a household with a high
level of income.
64 Y. C. Wong et al.

METHODOLOGY

Procedure
There have been various attempts at both the national and international level
to measure the digital divide within a society. According to the ITU (2005),
most of these measurements mainly compare the degree of access to ICT
among different demographic groups, usually disadvantaged groups (such
as the elderly and those with a low income) with more technologically
advanced groups (such as younger and better educated people). The com-
parisons usually take three forms:

1. Method A: Finding the absolute difference between the two groups (tech-
nologically disadvantaged and nondisadvantaged). For example, if the
nondisadvantaged group had an access rate of 90% compared with the
disadvantaged groups access rate of 45%, the difference between them
would be 45%.
2. Method B: Computing the ratio between the levels of access of the two
groups. In this case, the ratio is 0.5. Both Methods A and B can be used
to trace the changes across years to see whether the gap has narrowed
or enlarged. For example, if ICT access for the disadvantaged group
reaches 60% in five years and for the nondisadvantaged group 100%,
the difference is reduced to 40% by Method A (a reduction of 5%) and
0.6 by Method B (also an improvement).
3. Method C: Comparing the rates of change in the level of access between the
two types of group. For example, the rate of access for the nondisadvan-
taged group is 11.1% (from 90% to 100%), while that of the disadvantaged
group is 33.3% (from 45% to 60%) over a period of five years. However,
given that the disadvantaged groups usually have a smaller base, the rate
of growth will certainly be larger than that of the nondisadvantaged group.

The Digital Inclusion Index (DII) was constructed based on the concep-
tual framework regarding digital inclusion outlined earlier. It attempts to
compare the digital gap between disadvantaged groups and nondisadvan-
taged groups in the same society. To do this, it largely adopts Method B,
in which the levels of accessibility, knowledge, usage, and affordability of
ICT for various disadvantaged groups are compared with the nondisadvan-
taged group. Subsequently, it provides information about digital inclusion
of all groups and is able to show the relative position in terms of digital inclu-
sion among various disadvantaged groups and in what areas.
DII is computed for selected disadvantaged groups. In Hong Kong, the
groups selected were: (a) older people (older adults aged 60 or above), (b)
new arrivals (Hong Kong residents aged 10 or above who have resided in Hong
Kong for less than seveb years, (c) single parents (divorced=widowed=never
Digital Exclusion in an Information Society 65

married persons who live together with at least one child aged below age 18),
(d) female homemakers (women aged 3559 whose highest level of education
completed was primary six or lower), (e) young people aged 1017 in low
income households (with incomes lower than half of the median household
income), and (f) people with a disability or chronic illness (PWD=CI, with
functional impairment and=or diagnosed chronic illness).
Individuals with multiple disadvantaged characteristics were assigned to
all the relevant disadvantaged groups to compute the DII. Thus, information
about a person aged 60 or above with a chronic illness will be used in both
disadvantaged groups. However, when DII was computed for all disadvan-
taged persons considered together as one single group, this person was
counted only once to avoid double counting. Representative samples were
obtained from local households for each selected group.

Sampling
Three thousand, two hundred addresses of households were obtained from
the sampling frame maintained by the Census and Statistics Department,
Hong Kong Special Administrative Region (HSAR). Out of the 3,200
addresses, there were 2,682 valid addresses. Screening was conducted
through household visits to identify whether disadvantaged persons
described previously existed and in what number. All the identified people
were then invited to complete the questionnaire face-to-face by an inter-
viewer. A total of 1,893 households completed the screening (a success rate
of 70.6%). A total of 5,984 persons were identified, and among them 2,554
were disadvantaged persons. The interviewers successfully enumerated
2,312 of them. The enumeration results are presented in Table 1.
A representative sample of 284 persons belonging to the nondisadvan-
taged group was drawn from households containing no disadvantaged
persons (including the respondents). They formed the goalpost against
which to compare the accessibility, usage, knowledge, and affordability of
ICT for various disadvantaged groups.

Measures
Subindex scores were calculated for each pertinent dimension. As a result,
there are four subindexes, namely accessibility, knowledge, usage, and
affordability. The overall index score was the average of the four subindex
scores for each disadvantaged group. No weighting was applied.
Within each subindex, indicators were selected and each indicator given
an item score, which was expressed as a ratio between the value of the dis-
advantaged group for that indicator and the value of the comparison group
for that indicator. For example, in the accessibility subindex, one of the indi-
cators was the percentage of individuals having a home computer. The item
66 Y. C. Wong et al.

TABLE 1 Enumeration Results of Various Disadvantaged Groups from 1,893 Households

Individuals % of all the Individuals % interviewed


identified individuals (%) interviewed (%)

Older people (60 ) 1234 20.6 1107 89.7


New arrivals 326 5.4 296 90.8
Single parents 71 1.2 61 85.9
Female homemakers 465 7.8 452 97.2
PWD=CI 1033 17.3 966 93.5
Young people (low income) 312 5.2 277 88.8
All disadvantaged persons 2554a 42.7b 2312 90.5
a
Some individuals belong to different disadvantaged groups. The total is not the sum of the individuals in
each group.
b
For the same reason, the total percentage is not the sum of the percentages in each group.

score is the ratio between the percentage of persons having a home compu-
ter in the particular disadvantaged group under study and that of the percen-
tage of nondisadvantaged persons having a home computer. For example, if
the percentage value for the disadvantaged group is 30% and the comparison
group is 60%, the item score is 0.5. The subindex score was computed
through a combination of multiplication and summation of the related item
scores depending on the nature of the indicators.

ACCESSIBILITY SUBINDEX
Table 2 presents the five indicators and the formulae for computing the item
score for each indicator in the accessibility subindex. The ratios between the

TABLE 2 Accessibility Subindex

Goalpost:
Disadvantaged Non-disadvantaged Item
Indicators groups groups scores

1. % having a computer at home AD1 AG1 A1 AD1=AG1


2. % having a computer connected AD2 AG2 A2 AD2=AG2
to Internet at home (among
those having a home computer)
3. % that can use a computer AD3 AG3 A3 AD3=AG3
without permission (for those
who have used a computer and
have a home computer)
4. % having no constraints in using AD4 AG4 A4 AD4=AG4
software (for those who have
used a computer and have a
home computer)
5. % having broadband connection AD5 AG5 A5 AD5=AG5
(for those who have used a
computer and have a home
computer)
Digital Exclusion in an Information Society 67

percentages for the disadvantaged and the nondisadvantaged groups form


the item scores for each indicator. The first item, Percentage having a com-
puter at home is the core indicator, while the remaining four are discounting
factors. The discounting factors are limitations encountered by individuals in
terms of access. The overall score for the accessibility subindex is the dis-
counted core scores. In mathematical terms, this can be expressed as the
core score multiplied by the discounting factor scores:

Accessibility Subindex score A1  A3  A4 A1  A2  A5=2

Indicator items 3 and 4 are discounting factors for computer usage,


while items 2 and 5 are discounting factors regarding Internet access.

KNOWLEDGE SUBINDEX
The knowledge subindex measures the divide in knowledge about compu-
ters and the Internet between the disadvantaged and nondisadvantaged
groups. The nine indicators in this subindex include the percentage of per-
sons having medium or above levels of knowledge in using computers,
the Internet in general, and various Internet services in particular (e.g.,
searching for information, sending and receiving e-mail, joining chat rooms
and forums). It also includes the proportion of people who have taken com-
puter training courses. Table 3 presents the indicators and calculations for the
knowledge subindex. The overall score is the average of these item scores. It
is expressed mathematically as:

Knowledge subindex score K1 K2 K3 K4 K5 K6 K7 K8 K9=9

USAGE SUBINDEX
The usage subindex measures the gap between the usage of computer and
Internet services among the disadvantaged and nondisadvantaged groups.
Conceptually, the starting point for comparison is the percentage of people
who have used a computer and Internet services. However, in order to have
a more accurate picture of the level of usage, we have also to consider the
frequency and amount of time spent in using them as well as the types of
Internet services used. Thus, in calculating the usage subindex score, compu-
ter and Internet usage is discounted by the duration and frequency of
usage and types of service being used.
Table 4 presents the indicators for comparison; indicator 1 and 2 are the
percentage of computer and Internet usage while indicator 3 and 4 measure
the frequency (percentage of using Internet daily) and duration (average
hours online each time). Indicators 5 to 10 measure the percentage of the
use of six types of Internet services in the past three months. Indicator 3
and 4 are considered together as one discount factor relating to frequency
68 Y. C. Wong et al.

TABLE 3 Knowledge Subindex

Goalpost:
Disadvantaged Non-disadvantaged Item
Indicators groups groups scores

1. % having medium level or above KD1 KG1 K1 KD1=KG1


of knowledge in using
computer
2. % having medium level or above KD2 KG2 K2 KD2=KG2
of knowledge in using Internet
3. % who know in general=know KD3 KG3 K3 KD3=KG3
well how to search for
information
4. % who know in general=know KD4 KG4 K4 KD4=KG4
well how to send e-mail
5. % who know in general=know KD5 KG5 K5 KD5=KG5
well how to join chat rooms,
forums
6. % who know in general=know KD6 KG6 K6 KD6=KG6
well how to shop online
7. % who know in general=know KD7 KG7 K7 KD7=KG7
well how to get e-service
8. % who know in general=know KD8 KG8 K8 KD8=KG8
well how to publish a web page
9. % who have taken a computer KD9 KG9 K9 KD9=KG9
training course

and duration, while indictors 5 to 10 form another discount factor relating


to the use of Internet services. As with the other subindexes, the item score
for each indicator is expressed as a ratio between the values for the disadvan-
taged and nondisadvantaged groups. In mathematical terms, the formula is
expressed as:

Usage subindex score U1  U2  U3 U4=2 U5 U6 U7


U8 U9 U10=6=2

AFFORDABILITY SUBINDEX
The affordability subindex measures the gap in affordability between the
disadvantaged and the nondisadvantaged groups in purchasing equipment
and subscribing to services for computer usage and Internet access. Concep-
tually, a line must be drawn to delineate the amount of household income
that separates those who are considered able to afford the necessary equip-
ment and services for computer and Internet use and those who are not. It is
difficult to drawn this line, because even for households with a very tight
budget (such as those close to the local social assistant payment level) can
still squeeze some money (though with undesirable effects on their mem-
bers physical health and social life) to buy very basic facilities if it is a very
Digital Exclusion in an Information Society 69

TABLE 4 Usage Subindex

Goalpost:
Disadvantaged Non-disadvantaged Item
Indicators groups groups scores

1. % having ever used a computer UD1 UG1 U1 UD1=UG1


2. % having used Internet UD2 UG2 U2 UD2=UG2
3. % using Internet daily (of those UD3 UG3 U3 UD3=UG3
who have used a computer)
4. Average hours online each UD4 UG4 U4 UD4=UG4
time
5. % who have searched for UD5 UG5 U5 UD5=UG5
information (of those who
have used Internet in the last
3 months)
6. % who have sent e-mail, etc. UD6 UG6 U6 UD6=UG6
(those who have used
Internet in the last 3 months)
7. % who have used a chat room, UD7 UG7 U7 UD7=UG7
instant messaging, forum (of
those who have used
Internet in the last 3 months)
8. % who have shopped online UD8 UG8 U8 UD8=UG8
(of those who have used
Internet in the last 3 months)
9. % who have used e-service UD9 UG9 U9 UD9=UG9
(those who have used
Internet in the last 3 months)
10. % who have published a web UD10 UG10 U10 UD10=UG10
page (of those who have
used Internet in the last
3 months)

high priority for them. In our study, we chose to draw the line based on a
household income equal to or higher than half of the median household
income with respect to their household size. These households are
considered to have enough resources to buy computing equipment and sub-
scribe to Internet services without encountering great difficulties in other
aspects of life. Table 5 presents the single indicator for calculating this
subindex. Mathematically, it is expressed as:

Affordability subindex score AFD=AFU

Reliability and Validity of DII


The index is designed to measure digital inclusion of disadvantaged groups.
The unit of analysis is thus the group instead of individual. The computation
of the overall index and subindexes score, as illustrated previously, is based
on the combination of ratios of various indicators between the disadvantaged
70 Y. C. Wong et al.

TABLE 5 Affordability Subindex

Goalpost:
Disadvantaged Non-disadvantaged
Indicator groups groups Item score

1. % of people living in AFD AFU AF AFD=AFU


households with income
equal to or higher than
the median household
income for their
household size

groups and the nondisadvantaged. In examining the reliability and validity of


the index and subindexes, all the disadvantaged persons were randomly
assigned into 154 groups with 15 individuals each (the odd cases were not
included to make sure that all groups were of the same size). The items
and formulae of each subindex were then identified and the related item
scores computed. For the accessibility subindex, the formula is:

1=2A1  A3  A4 A1  A2  A5

The subindex score is the sum of two computed items:

1=2A1  A3  A4 and 1=2A1  A2  A5

For the usage subindex, the formula

U1  U2  U3 U4=2 U5 U6 U7 U8 U9 U10=6=2

is in fact, the sum of the following 8 computed items:

1=4U1  U2  U3; 1=4U1  U2  U4; 1=12U1  Us  U5;


1=12U1  U2  U5 . . . 1=12U1  U2  U10

The knowledge subindex is more straightforward. It is the average of


the nine related items: 1=9(K1), 1=9(K2) . . . 1=9 (K9). Since the affordability
subindex is a single item, reliability test is not applicable.
The results of the reliability tests indicated that Cronbachs a for the
accessibility subindex was 0.834, for the usage subindex was 0.791, and
for the knowledge subindex was 0.929. The Cronbachs a for the whole
index was 0.743. All of them were well within the acceptable level.
Construct validity was examined by exploratory factor analysis with
varimax rotation. A five-factor solution was identified. The first factor con-
sisted of most of the computed items related to the more common indicators
such as searching information, sending e-mail, and using chat room in the
Digital Exclusion in an Information Society 71

usage and knowledge subindexes (eigenvalue, EV 5.67); the second factor


consisted most of the computed items in the knowledge subindex
(EV 3.715); the third factor consisted of the two computed items of the
accessibility subindex and the single item of the affordability subindex
(EV 2.072); the fourth factor consisted of the two items related to web-pub-
lishing usage and knowledge (EV 1.685). The final factor (EV 1.37)
related to two relatively rare usage items (shopping online and use of
e-services). The results indicated that the responses to the computed index
items by and large reflected the proposed conceptual framework of digital
inclusion. It also indicated that among the local disadvantage population,
accessibility was still very much related to affordability, and some of the rarer
ICT usage and knowledge are limited to a small subpopulation and could be
taken away from the index for the sake of parsimony.
The item related to receiving computer training in the knowledge sub-
index did not load into any factor, and if it was taken away from the subindex
the reliability score increased to 0.929, but the overall index score remained
the same. If both the rarer knowledge and usage items and the item related to
receiving computer training were deleted, the overall index score for all the
disadvantage groups changed slightly from 0.41 to 0.40 (Table 7, discussed in
the next section, presents the overall index score and subindex scores of the
full index).

FINDINGS

Table 6 present the percentages=values with respect to the various indicators


of the sub-indexes of DII. These figures were used to compute the item scores
and subsequently the sub-indexes and the overall DII scores. Table 7 presents
the item scores, subindex scores, and the overall DII scores. Each disadvan-
taged group as well as all disadvantaged persons had their own scores. If
the item score for any disadvantaged group is 1, this means that it has the same
level on that indicator as the nondisadvantaged. If the score is 0, it means that
the value of the indicator is 0 for the disadvantaged group.
In terms of the accessibility gap, the subindex score indicated that for
all of the disadvantaged persons in this study, the score was 0.62. This
means that the level of accessibility for disadvantaged people is only about
two-thirds that of the nondisadvantaged. Older people and PWD=CI had
the lowest scores (0.50 and 0.53 respectively), and had even lower than
the overall scores for all disadvantaged persons considered together
(0.62). However, the female homemakers (0.75), young people (low
income) (0.72), single parents (0.71), and new arrivals (0.70) all had scores
higher than the overall score for disadvantaged people. Although
female homemakers and single parents had much lower levels of
computer=Internet knowledge and usage, they were more likely to have
school-age children. In fact, the single parents in our study were those
TABLE 6 Indicators of the Digital Inclusion Index

Older New Single Female Young people All disadvantaged Non-disadvantaged


people arrivals parents home- makers (low income) PWD=CI persons persons

No. of Individuals 1107 296 61 452 277 966 2312 284


Indicators
Accessibility subindex
% having a computer at home 46.4 64.2 65.6 69.9 70.4 50.6 58.6 91.1
% having a computer connected to 92.8 93.7 95.0 94.3 94.9 92.4 93.9 97.8
Internet at home (for those having
a computer at home)
% who can use computer without 95.4 98.3 93.3 91.1 85.8 90.1 91.1 96.4
permission (for those who have
used a computer and have a home
computer)
% having no constraints in using 100.0 95.0 100.0 100.0 97.9 99.3 98.1 95.3
software (for those who have used
a computer and have a home
computer)

72
% who have a broadband connection 97.0 98.8 97.3 97.9 96.1 96.2 97.3 98.3
(for those with a home computer
connected to the Internet)
Knowledge subindex
% having a medium level or above 1.7 24.7 14.8 1.3 41.2 6.8 11.8 58.1
of knowledge in using computer
% having a medium level or above 1.6 24.3 13.1 2.0 41.9 7.2 11.9 57.8
of knowledge in using Internet
% know in general=know well about 4.4 42.2 24.6 5.1 68.6 11.7 20.4 69.7
how to search information
% know in general=know well about 3.5 39.2 24.6 3.5 64.6 10.6 18.8 68.5
how to send e-mail
% know in general=know well about 2.3 29.4 19.7 2.0 56.3 7.9 14.8 57.7
how to join chat room, forum
% know in general=know well about 1.3 14.5 14.8 1.5 27.8 5.7 8.3 38.8
how to shop online
% know in general=know well about 1.2 15.5 18.0 1.1 30.3 5.8 8.8 44.4
how to get e-service
% know in general=know well about 0.5 15.5 9.8 0.2 30.7 3.0 7.2 27.5
how to publish web page
% having taken computer training 0.5 8.1 11.5 1.3 14.4 2.0 3.8 10.7
course(s)
Indicators of the Digital Inclusion
Index (continued)
Usage subindex
% having ever used a computer 6.7 51.4 31.1 10.6 84.8 16.4 26.4 78.6
% having used Internet (among 85.1 94.7 84.2 83.3 96.6 90.5 93.3 99.0
computer users)
% who use Internet daily (those who 44.4 54.9 50.0 35.0 52.0 49.0 51.7 68.0
have used a computer)
Average hours online each time 1.26 2.11 1.86 1.78 2.31 2.48 2.23 2.35
% who have searched for 96.6 92.8 100.0 97.1 87.7 92.5 92.4 95.9
information (those who have used
Internet in the last 3 months)
% who have sent e-mail, etc. (those 50.0 71.2 56.3 42.9 64.1 71.4 66.3 81.2
who have used Internet in the last
3 months)

73
% who have used chat room, instant 10.3 46.8 18.8 20.0 51.8 27.1 39.8 44.3
messaging, forum (those who
have used Internet in the last 3
months)
% who have shopped online (those 1.7 2.2 0.0 5.7 2.7 17.3 5.9 8.4
who have used Internet in the last
3 months)
% who have used e-service (those 5.2 4.3 0.0 2.9 5.5 6.0 5.2 9.9
who have used Internet in the last
3 months)
% who have published web page 1.7 8.6 0.0 0.0 10.5 3.8 7.6 10.1
(those who have used Internet in
the last 3 months)
Affordability subindex
% of people living in households 47.7 68.6 45.8 59.7 0.0% 51.8 50.4 96.1
with equal to or higher than the
median household income for
their household size
74 Y. C. Wong et al.

TABLE 7 Digital Inclusion Index Scores

Young All
Older New Single Female people disadvantaged
Subindexes people arrivals parents homemakers (low income) PWD=CI personsa

Accessibility 0.50 0.70 0.71 0.75 0.72 0.53 0.62


Usage 0.04 0.50 0.19 0.06 0.88 0.17 0.26
Knowledge 0.04 0.51 0.42 0.05 0.91 0.14 0.25
Affordability 0.50 0.71 0.48 0.62 0.00 0.54 0.52
DII 0.27 0.60 0.45 0.37 0.63 0.34 0.41
a
The DII score for all the disadvantaged is calculated separately by including all disadvantaged persons in
the calculation instead of averaging the DII scores of various disadvantaged groups. This avoids double
counting of individuals with multiple disadvantaged statuses.

who had a child aged 17 or below at home and so were more likely to have
installed a home computer.
Since the unit of analysis of DII is the group, to test the statistical signifi-
cance of the DII scores across the groups the individuals in each disadvan-
taged group were assigned into units of 10 persons to compute their
group DII scores. The number of groups for single parents was only 6 and
was excluded from the statistical test. We ran the ANOVA tests for the
remaining five disadvantaged groups and the result F(4, 203) 129.7,
p < .001 indicated that the mean values of the DII scores of these disadvan-
taged groups were not all equal. Post-hoc analysis also indicated that the dif-
ferences in the scores presented in Table 7 were largely significant. However,
the differences between older people and homemakers and between female
homemakers and persons with disabilities=chronic illnesses had not reached
statistical significance.

DISCUSSION

In this study we have shown the importance and have outlined the frame-
work and procedures in developing an index to measure more accurately
the level of digital inclusion across various disadvantaged groups within a
society. The administration of the index involves household screening and
surveys to collect information from the designated disadvantaged groups
as well as from people in households with no disadvantaged individuals.
We have measured and presented the findings from data collected in Hong
Kong. This index goes beyond the existing measures of digital divide, which
typically involve using publicly available statistics regarding the computer
and Internet access and having experience in using the Internet. This index
captures more comprehensively the various dimensions related to the con-
struct. In addition, the data collected among the selected disadvantaged
groups provided the government and nongovernmental organizations
Digital Exclusion in an Information Society 75

(NGOs) useful and clear information about the level of digital inclusion in
various dimensions so that they can develop more targeted interventions.
The four subindex scores illustrate in what dimension(s) disparities in
opportunity exist. For example, the low DII scores for older people are not
so much about access; in fact, many of them have a computer at home but
did not use it. Access is also not primarily about affordability. The major pro-
blem lies in their usage and knowledge of ICT. The design of the computer is
extremely challenging for them because many have little knowledge of
English, and learning to use the input method for Chinese through an English
keyboard increases the difficulty further. Using a computer and the Internet is
not necessary for their everyday lives and thus does not appear to be worth
the effort required to partake in such a steep learning curve. Most of what they
need can still be satisfied without the use of ICTfor example reading news-
papers and phoning or meeting friends who are also mostly not ICT users.
Indeed, some may argue that such a life of physical and social activity for
elderly people is more healthy than remaining fixed in front of a computer.
On the other hand, knowledge and usage are not a big problem for
young people living in low-income households. However, since by defini-
tion all of them live in households with an income less than half of the med-
ian income for their respective family size, they would benefit from better
public access, or, better still, computer donation programs. The knowledge
level of single parents is better than that of some other disadvantaged groups,
but their usage levels are still very low, reflecting that even if they know how
to use ICT they are unable to take full advantage of it. This study did not
examine the specific barriers for single parents and further investigation is
needed. Children would certainly benefit if their parents had better knowl-
edge and appreciated the importance of ICT for their childrens learning. Par-
ents with little knowledge and experience of using ICT might worry
excessively about the undesirable aspects of using ICT, especially when they
do not know what their children are doing with the computer. They might
simply refuse or delay the installation of a computer at home.
People with a disability or chronic illness should benefit from the use of
ICT especially when mobility problems limit their access to information,
services, and social interaction. However, certain kinds of disability require
specific aids and adaptations in order for people to use ICTfor example,
the use of voice instead of text for persons with a visual impairment. Such
devices could be very expensive for individuals and thus outside financial
support would be needed.
The design of the DII captures the disparity in the accessibility, knowl-
edge, usage, and affordability of ICT between the disadvantaged and nondis-
advantaged groups within the same city. Thus, the results are free from
technological and cultural factors that can differ greatly between societies
or different parts of larger societies, such as urban and rural areas. The index
devised as part of this research is not culture specific, and, with some
76 Y. C. Wong et al.

modification, it could be used in any country to design policy=programs to


reduce the divide between the electronically connected and disconnected.
Similar data has been collected in Macau by another study team. With
information available from more cities in the future, our knowledge about
digital inclusion will be clearer, and learning from others strategies to pro-
mote digital inclusion will be more fruitful. The current study involved
household survey, which is accurate but expensive. Given the nature and
limited number of items in the index, data can be collected through tele-
phone survey, which can be more suitable for places where household tele-
phone numbers are readily available and can serve as a good sampling frame
for the targeted disadvantaged groups and general population.

Limitations of Study
The selection of items, to some extent, reflects local characteristics of Hong
Kong. For example, the issue of the disparity in the availability of ICT infra-
structure in the market is ignored because almost all residential districts in
Hong Kong are covered. The accessibility to public facilities is excluded because
public access almost fully covers the local neighborhood through networks of
NGOs, public libraries, government service units, some local schools, and even
certain shopping malls. It is only that it is less convenient to get access through
these facilities comparing to having home access, which is the case for overall
90% of the nondisadvantaged population. However, in cities where even the
more advanced sector of society is not certain about public access, comparing
both public and home access is important. This is crucial when DII is adopted in
countries or cities where the coverage of public access is limited.
International comparisons would require extra resources to collect the
additional household data that are not usually included in the regular house-
hold surveys, such as the reported level of ICT proficiency and usage of the
Internet for various purposes. Definition of disadvantaged groups might also
vary from country to country; for example women are excluded from educa-
tion in some countries but not in others. Nevertheless, it is possible to com-
pile a simplified version of the index based on more readily available
statisticssuch as the percentage of Internet usage, home computer penetra-
tion rates, and income levels across certain disadvantaged groups such as
older people, women, low-income householdsand to compare it with
the general population. Proxy measures, such as education levels, can also
be used instead of knowing how to use ICT. To verify this, we have used
the percentage of having home computer with Internet connection as the
proxy measure for accessibility, the percentage having used the Internet
for usage, and the percentage having completed secondary education or
above for knowledge while keeping the single item for affordability
unchanged. The result of these proxy measures indicated that the overall
index score increased slightly to 0.44 from 0.40 of the original score. While
Digital Exclusion in an Information Society 77

the accessibility subindex score was very much the same between the origi-
nal and proxy measures, the effects on usage and knowledge differed. For
example, when education attainment is used a proxy measure, the knowl-
edge subindex scores for older people and persons with disabilities=chronic
illnesses increased. Apparently, other factors mediate between education
level and computer knowledge. On the other hand, using education attain-
ment as the proxy measure reduced the knowledge subindex scores of the
young people because many of them are still yet to complete their secondary
education, but they are more knowledgeable in using the computer and
Internet than their education level suggests. Proxy measures appear to have
a good potential to serve as substitutes for items in the more comprehensive
index, but treatment of the knowledge and usage scores need to be worked
out for different disadvantaged groups to increase its validity.

CONCLUSION

The new information society has created fundamental changes in our lives. It
has enabled many (including individuals, countries, and societies) to derive
great benefit from it, but it has also magnified existing divisions and exclu-
sions as well as created new ones. To bridge the digital divide, or promote
digital inclusion, the first step is to have a framework and method to measure
what and where they are. This paper presents attempts to measure and moni-
tor the level of digital inclusion in one large city. Knowledge is the first step
to evaluate, compare, and address this new form of exclusion.

REFERENCES

Census and Statistics Department. (2007). Hong Kong as a knowledge-based economy:


A statistical perspective. Hong Kong: Hong Kong Special Administrative Region.
China Internet Network Information Centre. (2000). 6th statistical survey report on
the Internet development in China (July 2000). Retrieved November 6, 2005,
from http://www.cnnic.net.cn/download/2005/2005072601.pdf.
China Internet Network Information Centre. (2007). 20th statistical survey report on
the Internet development in china (July 2007, in Chinese). Retrieved on August
4, 2007, from http://www.cnnic.net.cn/uploadfiles/pdf/2007/7/18/113918.pdf.
Employment and Social Policy Council. (2000, November 20). Objectives in the fight
against poverty and social exclusion. Brussels: European Council. Retrieved from
http://ec.europa.eu/employment_social/social_inclusion/docs/approb_en.pdf.
European Commission. (2005). eInclusion Revisited: The local dimension of the
information society. Brussels: Author.
International Telecommunication Union. (2003). World telecommunication
development report 2003: Access indicators for the information society (7th
ed.). Retrieved on November 3, 2005, from http://www.itu.int/ITU-D/ict/
publications/wtdr_03/index.html.
78 Y. C. Wong et al.

International Telecommunication Union. (2005). Building digital bridges. Geneva:


Author.
Internet World Stats. (2007a). Broadband Internet statistics: Top world countries with
highest Internet broadband subscribers in 2007. Retrieved August 4, 2007, from
http://www.internetworldstats.com/dsl.htm.
Internet World Stats. (2007b). The Internet big picture: World Internet users and
population stats. Retrieved August 4, 2007, from http://www.internetworldstats.
com/stats.htm.
Moller, C., & Amouroux, A. (Eds). (2007). Governing the Internet: Freedom and
regulation in the OSCE region. Retrieved August 5, 2007, from http://www.
osce.org/publications/rfm/2007/07/25667_918_en.pdf.
Organisation for Economic Co-operation and Development. (2001). Understanding
the digital divide. Paris: OECD publishing. Retrieved from http://www.oecd.
org/dataoecd/38/57/1888451.pdf.
Organisation for Economic Co-operation and Development. (2003). OECD science,
technology and industry scoreboard, 2003. Paris: OECD Publishing.
Organisation for Economic Co-operation and Development. (2004). OECD informa-
tion technology outlook, 2004. Paris: OECD Publishing.
Sciadas, G. (2002, October). Unveiling the digital divide. Ottawa: Statistics Canada.
Retrieved from http://www.statcan.gc.ca/pub/56f0004m/56f0004m2002007-eng.
pdf.
U.S. Department of Commerce. (2000). Falling through the Net: Toward digital
inclusion. Washington: Author.
U.S. Department of Commerce. (2004). A nation online: Entering the broadband
age. Retrieved November 1, 2005 from http://www.ntia.doc.gov/reports/anol/
index.html.
Wong, Y. C., Law, C. K., Fung, J. Y. C., & Lam, J. C. Y. (2006). New exclusion in infor-
mation society. Scientiae et Sapientiae Forum, 8(3), 3845.
World Summit on the Information Society. (2003). Declaration of principles building
the information society: A global challenge in the new millennium. Retrieved
November 1, 2005, http://www.itu.int/dms_pub/itu-s/md/03/wsis/doc/
S03-WSIS-DOC-0004!!MSW-E.doc.

You might also like