The 'Digital Divide' Among Financially Disadvantaged Families in Australia
First Monday

The 'Digital Divide' Among Financially Disadvantaged Families in Australia by Jennifer McLaren and Gianni Zappalà

Abstract
The 'Digital Divide' Among Financially Disadvantaged Families in Australia by Jennifer McLaren and Gianni Zappalà
Despite figures suggesting that Australia is a high consumer of information and communication technologies (ICT), it is well documented that the pattern of this consumption is not spread evenly across the population; a 'digital divide' exists. In general, research suggests that people from higher socioeconomic backgrounds have greater access to ICT compared to those from lower socioeconomic backgrounds. A less well-researched area is the factors that may influence ICT access and usage within certain demographic and socioeconomic groups. This paper presents new data on the access and usage of ICT (computers and the Internet) by 3,404 households and 6,874 children from financially disadvantaged backgrounds. Fifty-nine per cent of the sample had a home computer and just under one-third had the Internet connected at home. The most common location for accessing the Internet was at school. A striking finding was the strong association between the level of parental education and ICT access and use. Schools are important in closing or levelling the access gap, as most students use computers and the Internet at school. However, considering the importance of having home Internet access for children's educational performance, the fact that almost three-quarters of students in this study did not use the Internet at home is of concern, particularly given that almost half of a comparable Australian population have home Internet access. Finding ways to increase the home access of low-income families to the Internet should therefore remain a policy priority for all sectors aiming to bridge the digital divide. Policies aimed at bridging the digital divide should also ensure that programs provide appropriate parenting support and emphasise the educational importance of having home access to computers and the Internet.

Contents

Introduction
The 'digital divide' revisited
ICT Access and Usage in Australia
Background to the data
Profile of the Sample
Key Findings
Discussion and Conclusions

 


 

"Australia has another great dividing range. In the age of the information economy, modems - not mountains - separate the population" (Manktelow, 2001).

++++++++++

Introduction

The release of the 2001 Australian Census figures this year has led to renewed concern among commentators that the gap between the 'haves' and 'have nots' with respect to the access and usage of information and communication technologies (ICT), commonly referred to as the 'digital divide', is widening (Mathewson, 2002). The Census revealed that nationally, an average of 42 per cent of Australians had used a computer at home in the week preceding the Census. With respect to the Internet, only 19 per cent of Australians had been online at home in the week before the Census. While these figures do not tell us how many households have a computer or the Internet at home they nevertheless provide a reasonable proxy that reveals that the 'digital divide' is still a real part of the Australian landscape. These figures also seem to dampen some of the more optimistic views that were being expressed about the 'digital divide' as recently as a year ago. A paper one of us presented at a forum on the new economy, for instance, stated:

"... should we worry about it [the digital divide]? Is it not just a transitory phenomenon that will sort itself out in the medium to long-term? Like all new technologies, there will be some disparity of access, often due to cost initially, but as the costs of the technology become cheaper, it will be less of a problem. Evidence from the U.S. suggests that this may indeed be the case. Internet access among some disadvantaged groups that previously had low connection rates is now rising steadily. In Australia, the ABS projects that by the end of 2001 ... every second household in Australia will have home Internet access" [1].

While there is some evidence that an increasing number of people have access to ICT, this is occurring more slowly than predicted by some analysts. More importantly, the evidence confirms that the probability of households and children having home access to ICT is strongly related to socioeconomic status (SES), namely access increases with higher levels of SES. We know less, however, about the factors associated with home access and usage of ICT within certain SES groups.

This paper presents new data on the access and usage of ICT (computers and the Internet) by households and children from financially disadvantaged backgrounds. The next section outlines the concept of the 'digital divide' and its relationship to socioeconomic status. In particular, there are significant educational implications of not having home access to computers and the Internet for children from financially disadvantaged backgrounds. An overview of the nature and source of the data analysed in this paper is then presented. This is followed by a discussion of the key findings in terms of the factors associated with the ownership and access of ICT, and the frequency and location of ICT usage. The final section concludes by outlining some preliminary implications of the findings.

 

++++++++++

The 'digital divide' revisited

The 'digital divide' argument is well-known - namely, that the unequal access and usage of ICT across the population - is compounding disadvantage for some, because having access to ICT is becoming central to being able to fully participate in the economic, social, political and cultural spheres of society (Lee et al., 2002). The advent and increasing sophistication of ICT has changed, and will continue to change, the way in which businesses, governments, communities and individuals operate and interact with each other.

Some of the key spheres in which ICT is influencing participation (or lack thereof) in society include:

Economic participation

  • Enabling people to search and apply for employment opportunities; and,
  • Many jobs now involve having minimum levels of ICT competency as prerequisites.

Education & lifelong learning

  • Opportunities for lifelong learning, especially for people who have not had experience of the formal education sector, are more easily accessed through distance and e-learning programs;
  • Access to ICT is central for 'online schools' for children living in remote areas;
  • Studies show that students, teachers and parents feel that computers have a positive effect on learning (Ainley et al., 2000);
  • Recent research from the U.S. shows that the presence of computers and Internet at home are strongly and positively associated with the academic outcomes of school children, particularly children from disadvantaged backgrounds (Wilhelm et al., 2002);
  • Given the increasing use of ICT by students at school, there is a risk that teachers and schools operate on the assumption that all children have access to computers and the Internet at home, which may influence their expectations of students' work and their computer literacy at school (Mathewson 2002); and,
  • Different levels of ICT access, support and skills between private and public schools may further exacerbate public versus private school disparities.

Access to services

  • Many government services are being increasingly provided over the Internet, as are billing and banking services, which often offer discounts for paying or accessing services online. Using the Internet for these services not only saves time but is more cost-effective. A recent study found that most people (73 percent) who incorporated the Internet into their everyday lifestyle were able to reduce the time spent on errands by four hours per week, and many (40 percent) saved up to $A30 per week (Centre for International Economics, 2001).

Political participation & social inclusion

  • Given the fact that the Internet is able to transmit information efficiently across geographical boundaries, it has the capacity to reduce some of the disadvantage associated with living in distant and remote locations;
  • The Internet is becoming increasingly important for political participation and the democratic process, with several political movements or protests now occurring via e-mail campaigns. Similarly, most political parties and several political representatives now use the Internet as a key means of communication with the electorate and constituents (Curtin, 2001);
  • Many cultural/leisure activities now involve or benefit from access to the Internet. In fact, the Internet is also "promoting social inclusion of traditionally marginalised groups such as the elderly, disabled and women with children " through facilitating communication and access to support networks [2].

The unequal access to ICT not only affects the lives of individuals who happen to be on the wrong side of the divide, but society as a whole (Perri 6 with Jupp, 2001). A 'technology' gap will have:

  • Economic consequences
    Australia will have lower productivity if fewer people have the opportunity to exploit the benefits of ICT (Lee et al., 2002); and,
  • Social consequences
    Australia will be less cohesive if the 'new' or 'information' economy/society becomes the preserve of an exclusive minority (Zappalà et al., 2002).

Furthermore, while having access to the Internet can bring several benefits such as those listed above, more recently, commentators have pointed out that the 'digital divide' is more than just a simple division between those with access to the physical hardware of the new ICT and those without. The concept needs to also encompass the broader social environment within which technologies operate. As one recent critic of the 'digital divide' label has argued:

"[A]ccess to ICT is embedded in a complex array of factors encompassing physical, digital, human, and social resources and relationships. Content and language, literacy and education, and community and institutional structures must all be taken into account if meaningful access to new technologies is to be provided" [3].

A simple but useful concept that encapsulates this idea is what has been termed the 'ABCs of the digital divide' - Access, Basic Training and Content [4]. It recognises that the divide is not solely about physical access to ICT, but also ensuring that people have the requisite resources and skills to use the technology appropriately. The data in this paper shed most light on the access issue [5].

 

++++++++++

ICT Access and Usage in Australia

Before we move onto examining the data in more detail, it is useful to briefly review some of the key studies and surveys that have sought to identify the extent of ICT usage by individuals in Australia as well as the factors that may be driving the differential access. The findings from four recent studies are summarised in Table 1, although comparisons are difficult because of the different sample sizes and timeframes of each particular survey.

The most reliable of the four is the survey conducted by the Australian Bureau of Statistics (ABS) in November 2000 (ABS, 2000a). It showed that just over half (56 percent) of all households in Australia had a computer in their home, and just over one-third (37 percent) had home Internet access. These figures represented a sharp increase in Internet access, as 1998 estimates by the ABS suggested that only one in eight households were connected to the Internet. Furthermore, on the basis of trends at the time, the ABS projected that every second household in Australia (or 50 percent) would have home Internet access by the end of 2001.

 

Table 1: Recent Australian data on Household ICT Access.

Study
ABS (2000a)
National Office for the Information Economy (NOIE) (2002)
Casson et al. (2002)
Ericsson (2002)
as cited in Connors (2002)
Time of survey
November 2000
September 2001
2000-2001
May 2002
Sample
3,200 households
500 households
1,252 households
2,000 individuals
 
Percent with Computer
Percent with Internet
Percent with Computer
Percent with Internet
Percent with Computer
Percent with Internet
Percent with Computer
Percent with Internet
All households
56
37
64
52
 
44
76
68
With children
74
48
     
58
   
No children
46
32
     
36
   
Household income
               
$A0-49K
37
21
     
22-37
   
>$A50K
77
57
     
67
   

 

Table 1, suggests the ABS projection was accurate, as data collected in September 2001 by the National Office for the Information Economy (NOIE) estimated that almost two-thirds (64 percent) of Australian households owned or leased a computer, and just over half of all households (52 percent) were connected to the Internet (NOIE, 2002). The other two studies listed in Table 1, are less comparable as their samples were skewed towards people in capital cities in the case of Ericsson, and towards rural areas in the other (Casson et al., 2002). The most recent study, based on a sample of 2,000 individuals across five state capitals, conducted by Ericsson Australia, found that three-quarters of Australians have a PC at home and almost 70 percent have home Internet access (Connors, 2002). Overall, these surveys confirm that on a comparative basis, Australia ranks highly (third in the world) in adopting 'Information Economy enabling technologies' (NOIE, 2002; DITR, 2002).

Despite these figures that suggest Australia is a high consumer of ICT, it is well documented that the pattern of this consumption is not spread evenly across the population (Zappalà et al., 2002). In brief, the 'usual suspects' of socioeconomic disadvantage are involved in the digital divide:

  • Income: Level of income is an important factor in determining who benefits from the new technology. In 1998-1999, for instance only six percent of households on incomes less than $A19,000 were connected to the Internet compared to 47 percent of those on incomes of more than $A84,000 (Hellwig and Lloyd, 2000). In 2000 the disparity between income groups was still relatively high, with income earners in the top bracket 3.5 times more likely to have an Internet connection at home than those in the lowest bracket. The ABS survey found that households on incomes of $A50,000 or greater are twice as likely as households with incomes less than $A50,000 to have a home computer and Internet access (ABS, 2000a). A key reason why low-income households with computers do not have Internet access is due to the costs of connection (Curtin, 2001).
  • Level of education: The study by researchers at the National Centre for Social and Economic Modelling (NATSEM) found that, with all else being equal, educational attainment of an individual was a stronger predictor of having home computers and the Internet than income (Hellwig and Lloyd, 2000). Individuals with a university education were 2.5 times more likely to have home access to the Internet than those without.
  • Geographic location: Although the connection between the 'urban-rural divide' and the 'digital divide' is subject to debate, where a person lives does appear to influence their home access to the Internet. While the proportion of adults with Internet access at home in metropolitan areas grew from 24 to 30 percent between 1998 and 1999, the corresponding increase in non-metropolitan areas was from 15 to 18 percent (Hellwig and Lloyd, 2000). The latest figures from the ABS suggest that the gap between city and country in terms of Internet access is decreasing, with 40 percent of all metropolitan households having access compared to 32 percent of all households in non-metropolitan areas. Furthermore, once studies control for the influence of education and income, the influence of geographic location diminishes. This suggests that the observed differences between metropolitan and non-metropolitan areas is a function of the different socioeconomic characteristics of metropolitan and non-metropolitan populations, in particular, the lower income and qualification levels of the latter. As one researcher has stated, "Geography may not determine it [Internet access], but there is obviously a geographical dimension to it" [6].
  • Age: Adults aged over 55 are significantly less likely to have Internet access compared to younger groups in the population (ABS, 2000a).
  • Gender: The role of gender is unclear, with some studies finding that females have lower take-up rates for the Internet than males (ABS, 2000a) while other studies find that gender plays little to no role in access (NOIE, 2002).
  • Occupation: Blue-collar workers are less likely to be connected to the Internet at home compared to other occupational groups after controlling for income and qualifications. Those in lower income jobs are also less likely to use a computer or access the Internet at work (Hellwig and Lloyd, 2000).
  • Family type: Households with children are more likely to have home computers and Internet access compared to households without children. One-parent households, however, are far less likely to have access to the Internet (26 percent) than two-parent households (51 percent) (ABS, 2000c).
  • Indigenous status: Indigenous Australians are less likely to have home computers and Internet access compared to non-indigenous Australians.

Most of these findings confirm that people from higher socioeconomic backgrounds have greater access to ICT compared to those from lower socioeconomic backgrounds. Another important dimension is the factors that may influence ICT access and usage within certain demographic and socioeconomic groups. In particular, what factors are associated with home computer and Internet access for children from low socioeconomic backgrounds? There is little research that has specifically addressed this particular issue (see Funston and Morrison, 2000 for an exception).

 

++++++++++

Background to the data

The data for this section come from administrative records of students and families on The Smith Family's (TSF) Learning for Life (LFL) program [7]. The LFL program aims to increase the participation of children from financially disadvantaged families in the educational process by the provision of financial and educational support (see Zappalà and Parker, 2000; Smyth et al., 2002 for an overview of the program). As part of developments and enhancements to the program aimed at increasing access and usage of ICT by students, a small survey was included as part of the annual communication to families in October 2001. The main aim of the survey was to collect benchmark data on computer and Internet access and usage among LFL students. Although the survey was sent to parents in 5,850 households, they were asked to pass on the survey/s to their child/children to complete. Of the total students in the population (11,948), 7,226 completed the surveys, giving a response rate of 61 percent. Each survey contained a unique student code to enable responses to be matched to background information contained in TSF's Client Services Management Information System (CSMIS) database.

Following data entry and the matching of responses to the relevant background information, several steps were taken to clean the data and arrive at the two final samples used for this analysis. First, the 7,226 student responses were screened for internal inconsistencies. For instance, 352 cases were removed because the student had answered 'no' in response to the question 'Do you ever use the Internet?' but also answered 'sometimes', 'often' or 'regularly' to another question on how often they use the Internet. This left a final student database of 6,874 students.

Second, given that almost 85 percent of students had siblings who also took part in the survey, a database of 'households' that responded was created [8]. This was particularly important for examining the extent of household access to ICT. Responses to questions such as 'Do you have a working computer in your home?' for instance, would be misleading if analysed on an individual basis, since two siblings answering 'yes' to this item does not mean that there are two households with a computer. The 'household' database allows the level of analysis to be the 'family unit' rather than the individual student.

Third, creating a household database enabled us to further filter and screen the sample so that inconsistent responses between siblings from the same household could be removed [9]. This left a final sample of 3,404 households. This represents 58 percent of the total number of households that were on LFL at the time the data were collected. Fourth, as is discussed below, the respondents and non-respondents did not differ greatly in terms of the key characteristics.

 

++++++++++

Profile of the Sample

Table 2 presents the characteristics of the sample by a range of socio-demographic and socioeconomic characteristics:

  • Almost half (47 percent) of the students were in Years 7-10 with just under one-third in Years 4 to 6. There was no difference between respondents and non-respondents in terms of student age.
  • There was an even split between male and female students. Once again, there was no difference between respondents and non-respondents in terms of student sex.
  • Most of the students that responded (59 percent) lived in non-metropolitan areas [10]. Students who lived in metropolitan areas were slightly less likely to have responded (46 percent of non-respondents came from metropolitan areas compared to 41 percent of respondents).
  • Over two-thirds (68 percent) of students lived in areas that were below the median level of locational disadvantage in Australia as measured by the Index of Relative Socio-Economic Disadvantage (IRSED). The IRSED is one of five Socio-economic Indexes for Areas (SEIFA) derived from the 1996 Census of Population and Housing. The indexes relate to socio-economic aspects of geographical areas. The IRSED is derived from features such as low income, lack of English language fluency, low educational attainment and high unemployment. A low score on this index indicates that the area has high levels of low-income families and individuals in unskilled occupations with little training. The percentile indicates the relative extent of disadvantage compared with other communities in Australia. For example, living in an area that scored in the bottom decile indicates that the families in the area are on average worse off than 90 per cent of the rest of the families in Australia. An IRSED score was calculated for each case in the sample based on their post-code and then converted into percentile bands. There was no difference between respondents and non-respondents in terms of this indicator.
  • The majority (59 percent) of students that responded came from one-parent families. Students from one-parent families were also less likely to have responded (66 percent of non-respondents came from one-parent families).
  • Over two-thirds (69 percent) of the students have parents with ten or less years of education (i.e. Completed up to or less than Year 10). There was no difference between respondents and non-respondents in terms of level of parental education.
  • Approximately five out of every six students were from an English-speaking background. There was little difference between respondents and non-respondents in terms of ethnic and cultural background.
  • An overwhelming majority (90 percent) of the students came from households where social security was the main source of income. There was no difference between respondents and non-respondents in terms of this indicator (91 percent of non-respondents were also from households where social security was the main source of income).
  • Just under half of the students (44 percent) lived in public housing, just over one-third (36 percent) were from families that lived in privately-rented accommodation and one-fifth were from families that either owned or were paying off their own homes. Students who lived in public housing were less likely to have responded (51 percent of non-respondents), while those whose parents owned or were paying off their own homes were slightly more likely to have responded (20 percent compared to 13 percent of non-respondents).

 

Table 2: Socio-demographic characteristics of survey sample.

Student Characteristic
Number [11]
Percent
Year level at school
   
1-3
886
13
4-6
2,023
30
7-10
3,214
47
11-12
701
10
Total
6,824
100
     
Sex
   
Male
3,407
50
Female
3,461
50
Total
6,868
100
     
Location
   
Metropolitan
2,800
41
Non-metropolitan
4,074
59
Total
6,874
100
     
Level of locational disadvantage (IRSED) [12]
   
Bottom
1,273
19
10-25 percent
1,484
22
25-50 percent
1,874
27
50-75 percent
1,432
21
75-90 percent
570
8
Top 10 percent
195
3
Total
6,828
100
     
Family type
   
One-parent
3,933
59
Two-parent
2,787
41
Total
6,720
100
     
Parental education [13]
   
< Year 10
1,183
22
Year 10
2,592
47
Year 12
698
13
TAFE/Other post-secondary
608
11
University degree
378
7
Total
5,459
100
     
Ethnic/cultural background [14]
   
Anglo-Australian
5,348
79
Aboriginal/Torres Strait Islander (ATSI)
100
1
English speaking background (ESB)
201
3
Europe
332
5
Asia
138
2
Middle East & Africa
517
8
Central & South America
99
1
Pacific Islands
68
1
Total
6,803
100
     
Main source of income
   
Social security
5,980
90
Employment
630
10
Total
6,610
100
     
Housing type
   
Public rental
2,986
44
Private rental
2,388
36
Owns/purchasing house
1,356
20
Total
6,730
100

 

 

++++++++++

Key Findings

Home access to computers and the Internet

Overall, using the household sample, 59 percent of families had a computer at home. At first, this appears to be a higher level of ownership than that revealed by the ABS survey cited in Table 1. A more appropriate comparison, given that our sample comprises only households with school-aged children, would be with computer ownership among households with dependent children under the age of 18 who have access to a computer. This suggests that LFL families are significantly below the national average, as almost three-quarters (74 percent) of all Australian households with dependent children have a home computer.

 

Table 3: Home computer and Internet access (LFL Households).

 
Computer
Internet
 
Percentage
Number
Percentage
Number
Yes
59
2,006
32
1,085
No
41
1,398
68
2,319
Total
100
3,404
100
3,404

 

Table 3 also shows that just under one-third (32 percent) of families were connected to the Internet at home [15]. These results are not too dissimilar to the level of home access revealed by the 2000 ABS survey, where 37 percent of households had access to the Internet. However, a greater level of disparity is revealed through the more meaningful comparison with households with dependent children, since 48 percent of all Australian households with dependent children under the age of 18 had home Internet access. Furthermore, it is also below the 58 percent of households with children that had home Internet access revealed by the more recent CLC survey (see Table 1).

Given that our sample comprises households that are all financially disadvantaged, it is not surprising that we would find lower levels of home access to computers and the Internet compared to families in the wider population. The remainder of this section examines the extent to which certain socio-demographic and socioeconomic factors are associated with home access of computers and the Internet within this group of financially disadvantaged households.

ICT home access and socio-demographic variables

Table 4 shows the proportion of households that had computer and Internet access at home according to a number of socio-demographic variables. Several points stand out:

  • The geographic location of the household had no influence in terms of having a home computer; households in metropolitan areas were only slightly more likely to have Internet access compared to those in non-metropolitan areas. This finding may seem to go against the commonly held view that the 'digital divide' has a spatial dimension (Curtin, 2001). Studies that have used multivariate techniques in examining Internet access, however, have found that the influence of geography disappears once variables such as education level and income are controlled for (Hellwig and Lloyd, 2000). Given that this sample comprises only low-income households, these initial findings suggest that geographic location per se is not a significant influence with respect to access to ICT (see also Curtin, 2001 on this point) [16].
  •  

    Table 4: ICT home access and socio-demographic variables.

    Characteristic
    Computer (percentage)
    Internet (percentage)
    Overall distribution
    59
    32
    Location
       
    Metropolitan
    59
    34
    Non-metropolitan
    59
    30
    Ethnic Background
       
    Anglo-Australian
    58
    30
    ATSI
    25
    15
    ESB
    50
    30
    Europe
    71
    53
    Asia
    81
    43
    Middle East & Africa
    64
    42
    Central & South America
    66
    43
    Pacific Islands
    28
    8
    Family type
       
    One-parent
    55
    28
    Two-parent
    66
    39

     

  • The ethnic/cultural background of the household seems to be associated with levels of ICT access. While caution is needed with respect to some groups given the small cell sizes (see Table 2), Indigenous households were much less likely to have a computer or Internet access at home compared to other groups, with the exception of households from the 'Pacific Islands' background. Households where the parent/s were either Australian-born or born overseas but from English-speaking backgrounds had similar levels of computer and Internet access to the overall mean. In contrast, households from Non-English Speaking Backgrounds (NESB), especially European, had higher levels of computer and Internet access than the overall mean.
  • Finally, family structure seems to be associated with access levels, with one-parent households having lower levels of access to a home computer (55 percent) and the Internet (28 percent) compared to two-parent households (66 percent and 39 percent respectively).

ICT home access and socioeconomic variables

Table 5 shows the percentage of households with home computer and Internet access according to a range of socioeconomic variables. It suggests that all these variables were associated with the level of ICT access, although some variables appear to have a stronger association than others.

 

Table 5: ICT home access and socio-economic variables.

Characteristic
Computer (percentage)
Internet (percentage)
Overall distribution
59
32
Level of disadvantage
   
Bottom 10 percent
52
27
10-25 percent
59
30
25-50 percent
59
32
50-75 percent
60
33
75-90 percent
67
40
Top 10 percent
67
35
Housing type
   
Public rental
53
26
Private rental
58
33
Owns/purchasing
73
43
Main source of income
   
Social security
58
31
Employment
72
44
Parental education
   
< Year 10
43
18
Year 10
58
31
Year 12
68
42
TAFE/Other post-secondary
68
38
University degree
88
57

 

  • Not surprisingly, households that were located in the most disadvantaged areas, were less likely to have a home computer (52 percent) and home Internet access (27 percent), compared to households situated in the least disadvantaged areas (67 percent and 35 percent respectively).
  • In terms of the type of housing that families lived in, households that owned or were purchasing their homes were more likely to own a computer (73 percent) than households that were renting privately (58 percent) or living in public housing (53 percent). Owners/purchasers were also more likely to have Internet access (43 percent) compared to those renting privately (33 percent) or in public housing (26 percent).
  • A household's main source of income was also associated with home computer ownership and Internet access. Households whose main source of income was social security were far less likely to have computer at home compared to those whose main source of income came from employment (58 percent v. 72 percent). Similarly, home Internet access was higher for households whose primary income was from employment (44 percent) compared to those on social security (31 percent).
  • A striking finding was the strong association between the level of parental education and computer and Internet access. This is illustrated further in Figure 1. In households where the parent/s had less than ten years of education, 43 percent had a computer at home; this increased to 88 percent for households where the parent/s were university-educated. In households where parent/s had less than ten years of education, only 18 percent had Internet access at home; this increased to 57 percent with university-educated parent/s. This finding is consistent with previous studies that have found education level to be the key driver of Internet access, followed in importance only by income level (Hellwig and Lloyd, 2000).
  •  


    Figure 1: Households with home computers and Internet by education level.

     

    Frequency of usage of ICT by LFL students

    In addition to examining the ownership and access of ICT by households, we also examined how frequently students use computers and the Internet as well as where they access them.

    Computer Usage

    An overwhelming majority of students (98 percent) indicated that they used a computer. This is comparable to Australia-wide surveys that have found that 95 percent of children aged five to 14 used a computer in the last 12 months (ABS, 2000b). Of those that used a computer, Table 6 presents the variation in the frequency of usage by a range of demographic variables. Overall, most students stated that they use a computer 'sometimes' (33 percent) or 'often' (28 percent), with one quarter of students stating that they use a computer 'regularly'.

    Table 6 shows that student age is a key factor in discriminating among LFL students in terms of frequency of computer usage. The older students use computers more frequently than younger students. While none of the other demographic characteristics seem to be strongly associated with the frequency of computer use, boys were more likely to state they used a computer 'regularly' (28 percent) compared to girls (22 percent). In terms of ethnic/cultural background, students from a European background were most likely (32 percent) to use a computer 'regularly' compared to other groups; students from two-parent families were also more likely to use a computer 'regularly' (27 percent) compared to students from one-parent families (24 percent). Students from metropolitan areas were slightly more likely to use a computer 'regularly' (26 percent) compared to those from non-metropolitan areas (24 percent).

     

    Table 6: Frequency of computer use by demographic variables.

    Characteristic
    Frequency of computer use (percentage)
     
    Rarely
    Sometimes
    Often
    Regularly
    Number [17]
    Overall distribution
    14
    33
    28
    25
    6,694
    Year
             
    1-3
    18
    47
    20
    15
    846
    4-6
    16
    36
    27
    21
    1,981
    7-10
    12
    29
    32
    28
    3,131
    11-12
    13
    21
    30
    36
    690
    Sex
             
    Male
    14
    31
    27
    28
    3,310
    Female
    14
    34
    30
    22
    3,378
    Ethnic Background
             
    Anglo-Australian
    14
    33
    28
    25
    5,213
    ATSI
    12
    32
    29
    27
    96
    ESB
    10
    32
    32
    26
    196
    Europe
    12
    29
    26
    32
    325
    Asia
    10
    30
    33
    27
    133
    Middle East & Africa
    20
    31
    26
    23
    501
    Central & South America
    11
    36
    28
    25
    97
    Pacific Islands
    29
    29
    24
    18
    62
    Family type
             
    One-parent
    15
    33
    29
    24
    3,824
    Two-parent
    13
    33
    27
    27
    2,723
    Location
             
    Metropolitan
    15
    32
    27
    26
    2,727
    Non-metropolitan
    14
    33
    29
    24
    3,967

     

    Table 7 examines frequency of computer use by socioeconomic status. Once again, parental level of education seemed to have the most influence, with over one-third (35 percent) of students whose parents were university educated using a computer 'regularly' compared to 23 percent of students whose parents had not completed Year 10. Similarly, students whose parents' main source of income was from employment were more likely to state they used a computer regularly (29 percent), compared to students whose parents' main source of income was from social security (24 percent). Regular usage was also higher for students who lived in a house that was owned or being paid off compared to those in private or public rental accommodation and for those who lived in the more advantaged areas based on the IRSED.

     

    Table 7: Frequency of computer use and socioeconomic variables.

    Characteristic
    Frequency of computer use (percentage)
     
    Rarely
    Sometimes
    Often
    Regularly
    Number [17]
    Overall distribution
    14
    33
    28
    25
    6,694
    Parental education
             
    < Year 10
    18
    35
    25
    23
    1,136
    Year 10
    14
    34
    29
    24
    2,526
    Year 12
    16
    29
    27
    28
    688
    TAFE/Other post-secondary
    13
    34
    31
    22
    597
    University degree
    8
    25
    33
    35
    375
    Main source of income
             
    Social security
    15
    33
    28
    24
    5,815
    Employment
    11
    30
    31
    29
    620
    Housing type
             
    Public rental
    16
    34
    27
    23
    2,901
    Private rental
    14
    32
    29
    25
    2,319
    Owns/purchasing
    10
    30
    30
    29
    1,335
    Level of disadvantage
             
    Bottom 10 percent
    13
    33
    30
    25
    1,239
    10-25 percent
    15
    33
    29
    24
    1,446
    25-50 percent
    16
    33
    26
    26
    1,831
    50-75 percent
    16
    31
    30
    24
    1,382
    75-90 percent
    11
    33
    29
    27
    560
    Top 10 percent
    12
    33
    27
    29
    191

     

    Internet Usage

    Just over four-fifths of students (82 percent) indicated that they had used the Internet. Consistent with other studies (see ABS, 2000b), Figure 2 shows that older students were significantly more likely to state that they had used the Internet (95 percent for those in Years 11 and 12) compared to younger students (49 percent for those in Years 1 to 3).

     


    Figure 2: Internet use by age.

     

    Once again, the level of parental education was a key factor in the use of the Internet by students (Figure 3). While 92 percent of students whose parent/s were university educated had used the Internet, this fell to 76 percent for students whose parents had not completed Year 10. In terms of odds ratios, students whose parents completed Year 12 were one and a half times more likely to have stated that they had used the Internet than students whose parents did not complete Year 12. Those students whose parents had a university degree were almost three times more likely to have ever used the Internet than those whose parents did not have a university degree.

     


    Figure 3: Internet use by parental education level.

     

    Of those that used the Internet, Table 8 presents the variation in the frequency of usage by a range of demographic variables. Overall, only a small proportion of students stated that they used the Internet 'regularly' (11 percent), with just over one-fifth stating they used the Internet 'often' (22 percent), and almost two-thirds of students stating that they used the Internet either 'rarely' or 'sometimes'. Overall then, LFL students use the Internet less frequently than computers.

    Table 8 also suggests that student age is a key factor in discriminating between the frequency of Internet usage among LFL students, with older students using the Internet more frequently than younger students. Male students were more likely to state they used the Internet 'regularly' (13 percent) compared to female students (10 percent). In terms of ethnic/cultural background, students from a European background were most likely (17 percent) to use the Internet 'regularly' compared to other groups; there was little difference in Internet usage by students according to family structure. Students from metropolitan areas were slightly more likely to use the Internet 'regularly' (14 percent) compared to those from non-metropolitan areas (10 percent).

     

    Table 8: Frequency of Internet use and demographic variables.

    Characteristic
    Frequency of Internet use (percentage)
     
    Rarely
    Sometimes
    Often
    Regularly
    Number [17]
    Overall distribution
    32
    35
    22
    11
    5,565
    Year
             
    1-3
    44
    38
    13
    5
    423
    4-6
    36
    39
    18
    7
    1,606
    7-10
    28
    35
    24
    13
    2,861
    11-12
    29
    25
    27
    19
    658
    Sex
             
    Male
    31
    35
    22
    13
    2,731
    Female
    33
    35
    22
    10
    2,830
    Ethnic Background
             
    Anglo-Australian
    32
    36
    21
    11
    4,293
    ATSI
    33
    38
    21
    9
    77
    ESB
    31
    30
    23
    16
    172
    Europe
    24
    32
    27
    17
    287
    Asia
    38
    33
    18
    11
    118
    Middle East & Africa
    31
    31
    25
    13
    424
    Central & South America
    19
    38
    33
    10
    81
    Pacific Islands
    36
    36
    21
    7
    58
    Family type
             
    One-parent
    33
    36
    21
    11
    3,163
    Two-parent
    30
    35
    23
    12
    2,274
    Location
             
    Metropolitan
    30
    34
    23
    14
    2,301
    Non-metropolitan
    33
    36
    22
    10
    3,264

     

    Table 9 examines frequency of Internet use by socioeconomic status. In contrast to computer usage, there did not appear to be a strong relationship between socioeconomic variables and the regularity of Internet usage.

     

    Table 9: Frequency of Internet use and socioeconomic variables.

    Characteristic
    Frequency of computer use (percentage)
     
    Rarely
    Sometimes
    Often
    Regularly
    Number [17]
    Overall distribution
    32
    35
    22
    11
    5,565
    Parental education
             
    < Year 10
    33
    38
    18
    11
    894
    Year 10
    32
    36
    21
    11
    2,090
    Year 12
    32
    32
    24
    13
    587
    TAFE/Other post-secondary
    34
    33
    21
    12
    500
    University degree
    26
    35
    29
    10
    347
    Main source of income
             
    Social security
    32
    35
    22
    11
    4,819
    Employment
    32
    32
    25
    10
    536
    Housing type
             
    Public rental
    33
    36
    20
    12
    2,342
    Private rental
    32
    34
    23
    11
    1,945
    Owns/purchasing
    29
    36
    24
    12
    1,159
    Level of disadvantage
             
    Bottom 10 percent
    32
    34
    22
    12
    1,037
    10-25 percent
    32
    37
    21
    10
    1,204
    25-50 percent
    32
    35
    22
    12
    1,474
    50-75 percent
    31
    34
    23
    12
    1,167
    75-90 percent
    31
    35
    21
    13
    481
    Top 10 percent
    29
    35
    22
    14
    162

     

    Location of Internet use

    Almost three-quarters (70 percent) of students that used the Internet did so at school. Table 10 shows that the next most common location for Internet use was at home (29 percent). While the importance of school as a site for Internet use is consistent with other surveys in Australia, the proportion of students who indicated they used the Internet at home is lower compared to the national average. For instance, the ABS found that 67 percent of children aged between five and 14 used the Internet at school and 56 percent used the Internet at home (ABS, 2000b). Looking at a similar age group among the LFL students shows that while the same proportion (67 percent) was found to use the Internet at school, the rate for using the Internet at home was only 27 percent. Given the relatively low rates of home Internet access discussed earlier (32 percent), these findings are not that surprising, but more importantly, they suggest the important role that schools have as a means of providing access and training in ICT for students of disadvantaged backgrounds (Zappalà et al., 2002).

    It is also interesting to note that using the Internet at school was also related to the level of parental education. While two-thirds of students whose parents' had not completed Year 10 stated they used the Internet at school, this increased to almost four-fifths of students whose parents were university educated. Apart from the level of parental education, student age was the only other variable that influenced use of the Internet at school, with usage increasing for older students (38 percent of students in Years 1-3 used the Internet at school compared to 76 percent for students in Years 11-12).

     

    Table 10: Location of Internet Use.

    Place Internet used
    Number
    Percentage [18]
    During school
    4,790
    70
    Home
    2,024
    29
    Friend's House
    1,466
    21
    Public library
    1,113
    16
    At school after hours
    327
    5
    Youth/community centre
    90
    1
    Other
    252
    4
    Internet café
    66
    1

     

    ++++++++++

    Discussion and Conclusions

    This paper focused on what was termed the 'A' of the 'ABCs of the digital divide' - Access, Basic Training and Content [19]. Although the results have not employed multivariate techniques to isolate the effects of particular variables, they nevertheless point to several preliminary research and policy implications that will be pursued in more detail in forthcoming TSF publications and programs.

    First, while the access gap has been narrowing over the last few years, only one-third of families who were on the LFL program at the end of 2001 had home Internet access. This compares to almost half of the comparable (i.e. families with children) population Australia-wide. While some may not consider this finding to be that alarming, when seen in the context that having home Internet access is increasingly important for children's educational performance, then the fact that almost three-quarters of students did not use the Internet at home is of concern. Finding ways to increase the home access of low-income families to the Internet should therefore remain a policy priority for all sectors (government, private and nonprofit) aiming to bridge the digital divide.

    Second, the results are particularly interesting given that our sample controls for one of the key socioeconomic factors known to be associated with lack of access - income. All families on the LFL program are by definition low-income families. Despite this, several other dimensions of socioeconomic status seemed to be related to home access of computers and the Internet, and in some instances, the usage of computers and the Internet. In particular, the level of parental education was most strongly associated with home access to computers and Internet as well as computer and Internet usage. This finding is consistent with the key role found for educational level in home access to ICT in the multivariate analysis conducted by NATSEM (Hellwig and Lloyd, 2000).

    This finding also bears a remarkable similarity to other studies that examined the relationship between the educational performance of students on LFL and socioeconomic status (Zappalà and Considine, 2001; Considine and Zappalà, 2002). Controlling for other variables, the authors found that a student whose parent/s were university educated had a 39 percent predicted probability of attaining 'outstanding' results compared to nine percent for students whose parent/s had not completed Year 10. A key reason posited to explain that finding was that the levels of parental education acts as a proxy for the degree of educational support parents provide for their children. Previous studies show that the level of parental education is strongly associated with factors such as the home literacy environment, parents' teaching styles and investment in resources that promote learning (Shonkoff and Phillips, 2000). Key resources for learning in today's information society also include computers and the Internet.

    This has at least two implications. First, the costs of these resources, as with other educational costs in general, are increasingly being pushed onto individual families. This further compounds the problem for families in financial disadvantage who often struggle to meet the basic costs of their children's education. It therefore reinforces the need for programs such as Learning for Life that aim to assist families in financial disadvantage to meet some of the costs associated with their children's education. Second, policies aimed at bridging the digital divide should not only focus on reducing the cost of ICT, but also on ensuring that programs that provide appropriate parenting support also emphasise the educational importance of having home access to computers and the Internet. This may also mean that access and training programs should focus just as much on parents as they do on children. Once again, the dual-generation approach (focus on parents and children) of programs such as Learning for Life provide an appropriate framework within which to embed such initiatives.

    Third, other key factors associated with home access were ethnic and cultural background, family structure, housing type and regional disadvantage. The findings with respect to ethnicity were also consistent with the above-mentioned study on educational outcomes of LFL students. Namely, students from NESB (with the exception of those from Middle East/Africa) were significantly more likely to achieve outstanding results compared to students from English-speaking backgrounds. Similarly, the findings with respect to access suggest that families from some NESB groups have higher levels of home access compared to those that were either Australian born or born overseas from English speaking countries.

    Fourth, schools are important in closing or leveling the access gap, as most students use computers and the Internet at school. Reinforcing the role of parental education, however, the likelihood of students using the Internet at school also increased in line with the educational level of their parents. Greater research and policy attention needs to be given to the role of schools, teachers and parents in the 'ABC of the digital divide'. End of article

     

    About the Authors

    Jennifer McLaren is a Research Assistant in the Research & Social Policy Team, The Smith Family. Prior to joining the team in March 2002, she completed one year of a combined PhD/Masters in neuropsychology at Macquarie University, where she also worked as a research assistant.
    E-mail: Jennifer.McLaren@smithfamily.com.au

    Dr. Gianni Zappalà is Research Manager in the Research & Social Policy Team, The Smith Family. Prior to joining The Smith Family, he held various teaching and research positions at the Universities of Sydney, Cambridge (Fellow of Emmanuel College), Wollongong and the Research School of Social Sciences at the Australian National University. His research at TSF has included: the social impact of the new economy, the relationship between socioeconomic status and educational outcomes, volunteering, assessing the social impact of companies, and school-to-work transitions.
    E-mail: gianniz@smithfamily.com.au

     

    Acknowledgments

    We thank our colleagues Kirsten Buwalda, Martin Laverty, Ciara Smyth, and Mike Wilson for comments and assistance. We also thank Peter Huta from the National Office of the Information Economy (NOIE) for comments and suggestions on an earlier draft of this paper. The usual disclaimers apply.

     

    Notes

    1. Zappalà, 2001, p. 3.

    2. Robbins, 2000, p. 14.

    3. Warschauer, 2002, p. 6.

    4. Wilhelm et al., 2002, p. 2.

    5. See Smyth and Zappalà (2002) for an overview of TSF's Computer Clubs that while also aimed at providing access, have a focus on training and content.

    6. Curtin cited in Manktelow, 2001, p. 4; Curtin, 2001.

    7. Robbins, 2000, p. 14.

    8. Of the 6,874 students 5,818 were members of families that had more than one child on LFL.

    9. There were 114 households where the response of one sibling was inconsistent with that of another sibling for the question 'Do you have a working computer in your home?' This corresponded to 266 individual cases that were deleted from the database. There were 187 cases where the question 'Where do you use the Internet?' - 'At home' was endorsed by one sibling and not by the other. These cases were not deleted, as it is possible that one child used the Internet at home while their sibling did not.

    10. Geographic location coding was based on the household's post code and refers to the classification used by Australia Post; Capital city post codes are classed as Metropolitan and all other areas as Non-metropolitan.

    11. Total number of cases vary for each variable due to missing data.

    12. Percentiles indicate level of disadvantage relative to Australia as a whole. For example, 10-25 percent encompasses areas that are better off than at least 10 percent of Australia and at most 25 percent of Australia. An area falling in the 50-75 percent band is less disadvantaged than one falling in the 25-50 percent band.

    13. Education level of most highly educated parent.

    14. Refers to both first- and second-generation Australians.

    15. This figure was based on responses that endorsed the option 'At home' to the question 'Where do you use the Internet?' This proxy may underestimate the level of household access as there may be cases where a household had the Internet at home but the parent/s did not allow their child/children to use it.

    16. A possible reason for the apparent lack of a geographic location effect is the coding system used (see note [5]), which does not allow a sharper differentiation of the 'non-metropolitan' category. This category includes, for instance, all areas other than a capital city (e.g. cities such as Newcastle in New South Wales). This was one reason that post codes were linked to IRSED scores, thus providing another proxy for geographical location.

    17. Number of cases may vary for each variable due to missing cases.

    18. Does not add up to 100 percent because participants could endorse more that one option.

    19. Wilhelm et al., 2002, p. 2.

     

    References

    M. Ainley, V. Bourke, R. Chatfield, K. Hillman, and I. Watkins, 2000. Computers, Laptops and Tools. ACER Research Monograph, number 56. Melbourne: ACER Press.

    Australian Bureau of Statistics (ABS), 2000a. "Use of the Internet by householders, Australia," Catalogue number 8147.0. Canberra: ABS.

    Australian Bureau of Statistics, 2000b. "Children's participation in cultural and leisure activities, Australia," Catalogue number 4901.0. Canberra: ABS.

    Australian Bureau of Statistics, 2000c. "Household use of Information Technology, Australia," Catalogue number 8146.0. Canberra: ABS.

    Claudia Casson, Kate MacNeill and Derek Wilding, 2002. One size fits none: Australian telecommunications household profiles. Sydney: UNSW, Communications Law Centre.

    Centre for International Economics, 2001. Save@Home: Valuing the benefits of home Internet access. Canberra: National Office for the Information Economy.

    E. Connors, 2002. "It's follow my leader in gizmo game today," Australian Financial Review (29 May).

    Gillian Considine and Gianni Zappalà, 2002. "The influence of social and economic disadvantage in the academic performance of school students in Australia," Journal of Sociology, volume 38, number 2, pp. 129-148. http://dx.doi.org/10.1177/144078302128756543

    J. Curtin, 2001. "A digital divide in rural and regional Australia?" Current Issues Brief, number 1, 2001-02. Canberra: Department of the Parliamentary Library.

    Department of Industry Tourism and Resources (DITR), 2002. Australia as a Modern Economy: Some statistical indicators 2002. Canberra: DITR.

    A. Funston and M. Morrison, 2000. "Investing gaps and opportunities: young people's access to IT in Australia," Foundation for Young Australians.

    O. Hellwig and R. Lloyd, 2000. Sociodemographic barriers to utilisation and participation in telecommunications services and their regional distribution: A quantitative analysis. Canberra: NATSEM.

    L. Lee, D. Markotsis and T. Weir, 2002. "Social impacts of the New Economy," New Economy Issues Paper, number 5. Canberra: DITR.

    Nicole Manktelow, 2001. "The digital divide," Icon Magazine, Sydney Morning Herald (1-2 December), pp. 4-5.

    Catriona Mathewson, 2002. "Computer use figures reveal economic divide," Courier Mail (20 June), p. 12.

    National Office for the Information Economy (NOIE), 2002. The current state of play: Australia's scorecard. Canberra: NOIE.

    Perri 6 with Ben Jupp, 2001. Divided by Information? The 'digital divide' and the implications of the new meritocracy. London: Demos.

    Nick Robbins, 2000. "Internet," Socially Responsible Investment Position Paper. London: AMP Henderson Global Investors.

    Jack P. Shonkoff and Deborah A. Phillips (editors), 2000. From Neurons to Neighborhoods: The Science of Early Childhood Development. Washington D.C.: National Academy Press.

    Ciara Smyth and Gianni Zappalà, 2002. "The Smith Family Computer Clubs pilot program: A Progress report," Background Paper, number 6. Sydney: The Smith Family.

    Ciara Smyth, Gianni Zappalà and Gillian Considine, 2002. "Promoting participation and inclusion at school: A progress report on TSF's Learning for Life program," Briefing Paper, number 11. Sydney: The Smith Family.

    Mark Warschauer, 2002. "Reconceptualizing the Digital Divide," First Monday, volume 7, number 7 (July) at http://www.firstmonday.org/issues/issue7_7/warschauer, accessed 22 August 2002.

    Tony Wilhelm, Delia Carmen and Megan Reynolds, 2002. "Connecting kids to technology: Challenges and Opportunities," Kids Count Snapshot. Baltimore, Md.: Annie E. Casey Foundation.

    Gianni Zappalà, Vanessa Green and Ben Parker, 2002. "The disadvantaged in the new economy," In: Ron Callus and Russell Lansbury (editors). Working Futures: The Changing Nature of Employment Relations in Australia. Sydney: Federation Press.

    Gianni Zappalà, 2001. "On the hype and pitfalls of the new economy," invited keynote address at the Ideas at the Powerhouse: a festival of ideas, innovation and invention, Powerhouse Museum, Brisbane, 16-19 August, at http://www.ideasatthepowerhouse.com.au/5_highlights/paper_zappala.htm.

    Gianni Zappalà and Gillian Considine, 2001. "Educational performance among school students from financially disadvantaged backgrounds," Working Paper, number 4. Sydney: The Smith Family.

    Gianni Zappalà and Ben Parker, 2000. "The Smith Family's Learning for Life program a decade on: poverty and educational disadvantage," Background Paper, number 1. Sydney: The Smith Family.


    Editorial history

    Paper received 11 October 2002; accepted 25 October 2002.


    Contents Index

    Copyright ©2002, First Monday

    Copyright ©2002, Jennifer McLaren

    Copyright ©2002, Gianni Zappalà

    The 'Digital Divide' Among Financially Disadvantaged Families in Australia by Jennifer McLaren and Gianni Zappalà
    First Monday, volume 7, number 11 (November 2002),
    URL: http://firstmonday.org/issues/issue7_11/mclaren/index.html





    A Great Cities Initiative of the University of Illinois at Chicago University Library.

    © First Monday, 1995-2017. ISSN 1396-0466.