1. Introduction
The digital divide has become one of the great social concerns since the rapid permeation of ICTs into society that began in the mid 1990s. A considerable amount of policy-related, scholarly, or popular research and discussion surrounds this issue. This paper examines the development of the digital divide or social diffusion of different kinds of ICTs in Japanese society in the 1990s and 2000s, paying attention to the digital divide in other industrialized societies, especially, Korea, Finland and the United States. For, these societies have led social diffusion of ICTs and I would like to discuss relative characteristics of a Japanese way of the development or underdevelopment as an information society among major industrialized societies.
2. Emerging social concern about the digital divide
The Internet and mobile network began to grow rapidly in the middle of 1990s. Since then, the convergence of digital, networking and mobile technologies has been exploding. The rapid permeation of ICTs into society inspired not only visions of a utopian future, but also much unease about negative aspects. One of the great social concerns is the digital divide.
Though we do not know enough yet as to where and to what extent it will go, it is widely recognized that ICTs and their social diffusion have been changing the way people, things, capital, and information connect with one another. Not only industrial-economic but also political-administrative and socio-cultural domains undergo profound change and need reconfiguration. In addition, innovations of ICT related areas are highly dynamic. Those who adapt themselves to such dynamism are likely to enjoy the opportunities information networks could provide; those who fail to do so may be left behind.
Needless to say, it is widely acknowledged that it is a series of reports titled “Falling Through The Net," FTTN, issued by NTIA, National Telecommunications and Information Administration, the U.S. Department of Commerce, which played a critical role in drawing public attention to the issue of “the digital divide" in the late 1990s (NTIA 1995, 1998, 1999, 2000). While the 1995 report used the terms “haves" and “have nots," with regard to degree of access to advanced ICTs, i.e., personal computers, modems, and computer communications at the time, the report issued in July 1998 was subtitled "New Data on the digital divide," and dealt with access to the Internet. This report articulated, and stimulated great interest in the digital divide as an important social issue. The third report in 1999 indicated the considerable digital divides between ethnic, educational, and age groups, the sexes, and residential areas in the United States, and observed that the divide seemed to be widening, which stimulated public, scholarly and popular discussions over this issue.
Since then, a considerable amount of policy-related, scholarly, or popular research and discussion has occurred surrounding this issue, and various policy measures, initiatives and activities to overcome or bridge the digital divide have been set forth and carried out.
At the beginning of the discussion in the mid 1990s, the digital divide was generally defined as the gap between those who had and those who lacked access to the PC-based Internet. However, along with increases in the range of ICTs and their capabilities, it has become evident that "the digital divide" is not a matter of access to the PC-based Internet, but of different access means with a wide range of speeds and a great variety of use, as well as a matter of ICT skills, literacy and social support (see, Kling 1999, Hargittai 2002, 2003, Carvin 2000a, b, DiMaggio et al. 2004; Loader and Keeble 2004; Dewan and Riggins 2005).
Thus, DiMaggio and Hargittai (2001) suggest that the term "digital inequality” better captures the complexity of inequalities relevant to understanding the differences in access and use of information technologies. They set forth on five dimensions, along which digital inequality considers variation: technical means (software, hardware, connectivity quality); autonomy of use (location of access, freedom to use the medium for one"s preferred activities); use patterns (types of uses of the Internet); social support networks (availability of others one can turn to for assistance with use, size of networks to encourage use); and, skill (one"s ability to use the medium effectively).
3. Japanese society as an advanced and diversified information society and its digital divide
Given the diversification of ICTs and the development of conceptual frameworks for digital divide studies, the diffusion of ICTs in Japanese society in 1990s and 2000s provides an interesting and intriguing case. Table 1, 2 and 3 show the penetration rate of broadband internet connections and 3G mobile phones and internet-connectable mobile devices in various industrialized societies in the middle of 2000s. As is evident in these figures, the commercialization and social diffusion of FTTH and advanced mobile technologies in Japan, along with that in Korea, are astounding.
Table 1 OECD Broadband subscribers per 100 inhabitants, by technology, December 2006
|
DSL |
Cable |
Fiber/LAN |
Other |
Total |
Denmark |
19.6 |
9.4 |
2.6 |
0.4 |
31.9 |
Netherlands |
19.5 |
12.0 |
0.4 |
|
31.8 |
Korea |
11.4 |
10.7 |
7.0 |
|
29.1 |
Finland |
23.5 |
3.5 |
|
0.3 |
27.2 |
Sweden |
16.0 |
5.2 |
|
4.8 |
26.0 |
Canada |
11.4 |
12.3 |
|
0.1 |
23.8 |
United Kingdom |
16.5 |
5.1 |
|
|
21.6 |
France |
19.1 |
1.1 |
|
|
20.3 |
Japan |
11.1 |
2.8 |
6.2 |
|
20.2 |
United States |
8.5 |
10.3 |
0.3 |
0.6 |
19.6 |
Germany |
16.4 |
0.5 |
|
0.1 |
17.1 |
OECD average |
10.5 |
4.9 |
1.1 |
0.3 |
16.9 |
Source: Data from OECD ICT Key Indicators, http://www.oecd.org/dataoecd/24/32/38469057.xls; accessed October 30, 2009.
Table 2 3G subscribers per total mobile phone subscribers in G7+Korea and Finland, 2005
country |
Korea |
Japan |
Italy |
United Kingdom |
Finland |
Germany |
France |
United States |
Canada |
3G subscribers per total mobile phone subscribers (%) |
94.1 |
50.1 |
12.7 |
1.8 |
0.8 |
0.0 |
0.0 |
0.0 |
0.0 |
Source: Data adapted from OECD Communications Outlook 2007, Figure 4.6, p.98.
Table 3 Internet-connectable mobile devices as a share of total mobile phone subscribers in G7+ Korea and Finland, except for Canada, 2004
|
Japan |
Korea |
United States |
Finland |
Italy |
Germany |
United Kingdom |
France |
Internet-connectable devices (%) |
94.1 |
89.0 |
33.5 |
22.1 |
18.9 |
14.8 |
12.8 |
13.2 |
Source: Data from MIC Information and Communications in Japan 2005, Figure 5, p.86.
As of September 2006, the number of FTTH subscribers in Japan was 7.15 million. In Korea, almost 3 million people living in high-rise apartments subscribed to services providing FTTB, fiber to the building, or FTTC, fiber to the curb, coupled with high-speed LAN (local area network) connections. Japan and Korea are well ahead of Europe and the United States in these statistics. Both societies are also leaders in the spread of 3G mobile services. According to ENTER, a Spanish research agency, and IDATE, a European consulting firm, as of June 2006, the number of 3G customers is estimated to be just over 100 million in the world. The Asia-Pacific region accounts for 50%, 54 million, of which 29 million are in Japan and 11 million are in Korea.
However, Table 4 paints a different picture. The Koreans use the Internet most along with the Finns, but do not use mobile phones as much. The Japanese population does not use the Internet or mobile phone as much as other industrialized societies, in spite of its active use of advanced technologies.
Table 4 Mobile phone subscribers and Internet users per 100 inhabitants in G7+2, 2005
country |
Internet users per 100 inhabitants |
country |
mobile subscribers per 100 inhabitants |
Finland |
73.0 |
Italy |
123.1 |
Korea |
72.8 |
United Kingdom |
109.7 |
Canada |
67.9 |
Finland |
100.4 |
United States |
66.3 |
Germany |
95.8 |
United Kingdom |
53.8 |
France |
79.4 |
Japan |
51.7 |
Korea |
79.4 |
Italy |
48.2 |
Japan |
75.3 |
France |
43.2 |
United States |
71.5 |
Germany |
43.2 |
Canada |
51.4 |
Source: Data extracted from ITU Statistics Database, http://www.itu.int/ITU-D/ICTEYE/Indicators/Indicators.aspx, accessed January 10, 2010, except for the data of Internet users per 100 inhabitants of Japan, which is calculated by Kimura based on the data provided by MIC Information and Communications in Japan 2006, Figure 1-2-2, p.18.
Thus, there must be a great variety of digital divides in Japanese society. When it comes to hardware and physical connections to one or more digital networks, a considerable number of people are not connected to any network whatsoever, be it via PC or mobile phone, while as large a number of people subscribe to both an FTTH service and a state-of-the-art 3G mobile phone service. In between these two extremes, there are a variety of wired and wireless network access measures and devices. Again, as stated above, applications, services, and uses of information networks have become truly diversified.
Therefore, it is of great importance to investigate what kind of digital divide has been developing in Japan, if any. It is not sufficient to merely ask who has access to the PC-based Internet and who does not. Instead, we must first ask who uses what kind of ICT via what sort of access measures. Furthermore, we must investigate how diversified applications and services on different network platforms stimulate what kinds of use and what kinds of benefit through their use in Japan.
We begin with a scrutiny of the mobile divide. Table 5 summarizes the time-series data of the numbers of subscribers of fixed landline telecommunications service and various mobile services in Japan. Throughout the late 1990s, the number of mobile phone subscriptions grew so remarkably that it exceeded that of wired lines by the year 2000 (Table 5). It is not unusual for some, especially the young, to have a mobile phone alone, with no land line. Thus, the social diffusion of mobile phones and mobile internet is significant from a digital divide point of view.
Table 5 Time-series data of the number of subscriptions of different telecommunication services in Japan
March of each year |
Fixed Line |
Cellular |
Mobile internet |
3G |
1991 |
54,480 |
868 |
|
|
1992 |
56,210 |
1,378 |
|
|
1993 |
57,600 |
1,713 |
|
|
1994 |
58,780 |
2,131 |
|
|
1995 |
59,880 |
4,331 |
|
|
1996 |
61,040 |
10,204 |
|
|
1997 |
61,460 |
20,877 |
|
|
1998 |
60,380 |
31,527 |
|
|
1999 |
58,470 |
41,530 |
48 |
|
2000 |
55,440 |
51,139 |
7,499 |
|
2001 |
52,260 |
60,942 |
31,411 |
|
2002 |
51,000 |
69,121 |
50,336 |
89 |
2003 |
51,160 |
75,657 |
62,460 |
7,160 |
2004 |
51,590 |
81,520 |
69,732 |
16,690 |
2005 |
51,630 |
86,998 |
75,154 |
30,350 |
2006 |
50,563 |
91,792 |
79,758 |
48,330 |
2007 |
48,159 |
96,718 |
84,372 |
69,910 |
|
unit: thousand |
Source: Data extracted from MIC database, http://www.soumu.go.jp/johotsusintokei/, accessed October 30, 2009.
The MIC conducts the “Communications Usage Trend Survey” (CUTS) annually and provides rather detailed data and analysis, though they do not provide their raw data to the public. As for the use of cellular phones and mobile internet, rates are available by sociodemographic factors, i.e., age groups, sex, residential areas and household incomes. In certain years, the MIC applies multivariate statistical analysis to estimate the relative weight of these sociodemographic factors on the use rate.
Table 6 shows the results and analysis of the CUTS in December 2005. Among the sociodemographic factors analyzed, age is the strongest determinant factor for the use of both mobile phones and mobile internet. Elementary school children (parameter estimate is -1.92 for mobile phone use and -1.44 for mobile internet use) and the elderly aged 60 years and over (-1.04 for mobile phone use and -1.46 for mobile internet use) are the most unlikely to use mobile phones or mobile internet.
Table 6 Multivariate statistical estimation of the relative weight of sociodemographic factors on mobile phone use and mobile internet use among the Japanese as of December 2005
|
mobile phone use |
mobile internet use |
use rate (%) |
parameter estimate |
use rate (%) |
parameter estimate |
total |
71.9 |
|
58.2 |
|
sex |
male |
76.0 |
0.05 |
60.3 |
0.03 |
female |
68.1 |
-0.05 |
56.2 |
-0.03 |
age groups |
6 to 12 yrs old |
21.1 |
-1.92 |
22.5 |
-1.44 |
13 to 19 yrs old |
81.6 |
0.38 |
76.2 |
0.69 |
twenties |
96.6 |
0.88 |
84.8 |
0.9 |
thirties |
94.2 |
0.76 |
82.3 |
0.82 |
forties |
90.1 |
0.65 |
76.9 |
0.68 |
fifties |
80.9 |
0.28 |
56.6 |
-0.19 |
60 and over |
39.5 |
-1.04 |
19.1 |
-1.46 |
city size |
metropolitan areas |
75.5 |
0.05 |
62.3 |
0.07 |
cities |
70.7 |
-0.04 |
57.0 |
-0.02 |
towns and villages |
67.3 |
-0.01 |
51.8 |
-0.11 |
annual household income |
under \2 million ($17,241) |
53.4 |
-0.53 |
42.1 |
-0.39 |
\2 to 4 million ($34,483) |
64.7 |
-0.11 |
47.0 |
-0.18 |
\4 to 6 million ($51,724) |
71.9 |
0.00 |
57.5 |
-0.06 |
\6 to 8 million ($68,966) |
70.8 |
-0.04 |
59.9 |
0.02 |
\8 to 10 million ($86,207) |
77.4 |
0.15 |
63.6 |
0.13 |
\10 million and over |
79.3 |
0.10 |
66.2 |
0.15 |
sex |
|
0.10 |
|
0.05 |
age groups |
|
2.80 |
|
2.37 |
city size |
|
0.04 |
|
0.19 |
annual household income |
|
0.68 |
|
0.54 |
Source: Data and analysis extracted from MIC 2006.
Annual household income also has to do with the use of mobile phone or mobile internet. It is considerably less widely spread among members of those households with an annual income under \4 million ($34,483) than those with \4 million and over. On the other hand, members of those households with an annual income of \8 million ($68,966) and over are likely to enjoy both mobile phones and mobile internet.
Analysis of the CUTS is restricted to these four sociodemographic factors: age group, sex, residential area, and household income. However, education and occupation also might well be independent variables that influence the use of mobile phones and mobile internet.
In the following, I analyze the data from the JFK Survey. While this survey was conducted at the same time as the CUTS in 2005, the JFK Survey in Japan was restricted to Tokyo’s 23 wards. In addition, the respondents of CUTS were aged 6 and older while those of the JFK Survey were aged 20 to 69. The frequency distribution for mobile phones and mobile internet by sex, age, annual household income, education and occupation is summarized in Table 7. Then, nominal logistic regression analysis was applied to find out what factors are significantly correlated with their use.
Table 7 Frequency distribution for mobile phones and mobile internet by socio-demographic factors among the Japanese based on the JFK Survey
|
mobile phone |
mobile internet |
total |
89.7 |
60.7 |
sex |
male |
89.6 |
60.9 |
female |
89.8 |
60.4 |
age |
20s |
98.9 |
85.6 |
30s |
97.6 |
78.2 |
40s |
92.8 |
66.0 |
50s |
85.4 |
36.6 |
60 and over |
61.3 |
12.9 |
annual household income |
under \2 million ($17,241) |
74.1 |
51.9 |
\2 to 4 million ($34,483) |
83.5 |
57.0 |
\4 to 6 million ($51,724) |
85.6 |
55.7 |
\6 to 8 million ($68,966) |
94.6 |
62.2 |
\8 to 10 million ($86,207) |
96.1 |
64.7 |
\10 to 12 million ($103,448) |
97.1 |
73.5 |
\12 to 14 million ($120,690) |
94.4 |
66.7 |
\14 million and over |
100.0 |
62.5 |
education |
elementary or junior high school |
47.4 |
36.8 |
high school |
86.5 |
49.3 |
junior college or technical college |
97.7 |
69.3 |
college and graduate |
93.7 |
69.3 |
occupation |
full-time worker |
94.0 |
65.7 |
part-time worker |
91.0 |
57.7 |
student |
100.0 |
82.4 |
housewife |
79.4 |
50.8 |
no occupation (e.g., unemployed and retired) |
60.7 |
28.6 |
Data: JFK Survey
The results (Table 8) show that age is by far the strongest factor in determining the use of mobile phones and mobile internet. Annual household income (p=0.012) and education (p=0.042) also influences the use of mobile phones, while occupation has little to do with it.
In summary, in 2005, when mobile phones began to be widely used and their diffusion was approaching saturation, there is no significant difference in the use of mobile phones and mobile internet with regard to sex, occupation, or the size of a city’s population. While their use was differentiated most significantly by age, annual household income and education were also significant factors to some extent. These three factors show a greater influence of the use of mobile phones as a means of access to basic infrastructure in contemporary industrialized society than on the use of mobile internet, which can be considered a type of advanced service.
Table 8 Multivariate statistical estimation of the relative weight of sociodemographic factors on mobile phone use and mobile internet use among the Japanese based on the JFK Survey
|
mobile phone use |
mobile internet use |
ChiSq |
Prob>ChiSq |
ChiSq |
Prob>ChiSq |
sex |
1.023 |
0.3117 |
0.223 |
0.6369 |
age groups |
30.262 |
<.0001 |
104.592 |
<.0001 |
annual household income |
19.627 |
0.0118 |
8.417 |
0.3939 |
education |
9.741 |
0.0209 |
3.333 |
0.3431 |
occupation |
9.155 |
0.1030 |
5.318 |
0.3783 |
Data: JFK Survey
The CUTS in 2005 also provides the data of the use of a PC, internet (via any access means including mobile devices), the PC-based Internet and broadband (among the PC-based Internet users) by sociodemographic factors, i.e., age groups, sex, residential areas and household incomes and multivariate statistical analysis to estimate the relative weight of these sociodemographic factors. Interestingly enough, such an analysis was not provided for the use of the PC-based Internet.
Definitions of broadband Internet are diverse. Various minimum bandwidth values have been used in definitions of broadband. OECD defined broadband as having download data transfer rates equal to or faster than 256 kbit/s in its 2006 report on broadband. FCC’s definition has been data transmission speeds exceeding 200 kbit/s in at least one direction.
In addition, use of broadband is hard to define using social surveys. It is impossible for most respondents to know the speed of access; many users do not know what access measure they themselves use. In the CUTS, those respondents who answer DSL, cable modems, FTTH or FWA, fixed wireless access, are counted as broadband users. However, the number of FWA users is small enough to be negligible.
The number of households that answered the Internet was used at home was 2,716, which comprised 68.2% of the total of 3,982 households in the 2005 survey. Out of 2,716 households, 0.5% answered they used FWA as their means of access to the Internet at home. As far as the social surveys in which I got involved are concerned, the respondents who indicate that they use DSL, cable modems or FTTH are considered to be broadband users.
The frequency distribution of PC use, internet use (via any access means), broadband use and the PC-based Internet and the result of multivariate analysis the CUTS provides are put in Table 9. The pattern of relative weight among the three sociodemographic factors on these uses is almost the same as on the use of mobile phones. Age is the strongest; those aged 6 to 12 years and those aged 60 years and over are most unlikely to use PC, internet or broadband. Annual household income also influences their use to a certain extent; broadband use in particular appears to be fairly correlated with annual household income. Broadband is considerably less widespread among members of those households with their annual income under \4 million ($34,483). Sex and city size are not so strongly correlated with their use rate as annual income.
Table 9 Frequency distribution of PC, Internet, broadband use by sociodemographic factors and their multivariate analysis estimation of relative weights based on CUTS 2005
|
PC use |
Internet use (via any access means) |
Broadband use at home (among PC-based Internet users) |
PC-based Internet use |
|
rate (%) |
Para-
meter estimate |
rate (%) |
Para-
meter estimate |
rate (%) |
Para-
meter estimate |
rate (%) |
total |
56.7 |
|
69.1 |
|
37.3 |
|
52.3 |
sex |
male |
64.7 |
0.21 |
74.3 |
0.11 |
41.4 |
0.18 |
57.3 |
female |
49.1 |
-0.21 |
64.2 |
-0.11 |
33.4 |
-0.18 |
47.6 |
age groups |
6 to 12 yrs old |
58.5 |
-0.23 |
60.5 |
-0.63 |
37.6 |
-0.19 |
48.6 |
13 to 19 yrs old |
78.0 |
0.64 |
91.3 |
0.71 |
50.0 |
0.55 |
71.7 |
twenties |
78.3 |
0.74 |
92.3 |
0.75 |
51.5 |
0.64 |
72.4 |
thirties |
75.8 |
0.64 |
90.0 |
0.68 |
49.3 |
0.61 |
67.8 |
forties |
71.2 |
0.41 |
88.2 |
0.63 |
46.9 |
0.46 |
64.7 |
fifties |
55.0 |
-0.38 |
70.2 |
-0.23 |
33.7 |
-0.47 |
48.3 |
60 and over |
17.8 |
-1.82 |
26.5 |
-1.91 |
10.2 |
-1.59 |
15.7 |
city size |
metropolitan areas |
62.5 |
0.12 |
74.8 |
0.09 |
43.3 |
0.26 |
58.7 |
cities |
55.0 |
-0.03 |
67.4 |
-0.02 |
35.2 |
0.09 |
49.9 |
towns and villages |
48.1 |
-0.21 |
60.8 |
-0.17 |
29.4 |
-0.36 |
44.6 |
annual house-
hold income |
under \2 million ($17,241) |
34.4 |
-0.85 |
52.2 |
-0.54 |
21.5 |
-0.78 |
30.9 |
\2 to 4 million ($34,483) |
40.3 |
-0.63 |
54.7 |
-0.36 |
22.7 |
-0.77 |
33.4 |
\4 to 6 million ($51,724) |
55.2 |
-0.08 |
68.3 |
-0.15 |
33.9 |
-0.15 |
48.2 |
\6 to 8 million ($68,966) |
60.7 |
0.14 |
72.9 |
0.1 |
38.2 |
-0.04 |
55.3 |
\8 to 10 million ($86,207) |
63.8 |
0.21 |
75.0 |
0.17 |
45.6 |
0.38 |
60.5 |
\10 million and over |
68.5 |
0.43 |
78.4 |
0.29 |
49.7 |
0.69 |
67.6 |
age |
|
2.57 |
|
2.66 |
|
2.23 |
|
sex |
0.41 |
0.22 |
0.37 |
city size |
0.33 |
0.26 |
0.62 |
annual household income |
1.28 |
0.84 |
1.47 |
Source: Data and analysis extracted from MIC 2006.
And yet, compared with the use of mobile phone and mobile internet, PC use, internet use (via any access means) and broadband use are stratified more broadly by each of these sociodemographic factors. For example, by city sizes, the frequency distribution of mobile internet is 62.3 in metropolitan areas, 57.0 in cities and 51.8 in towns and villages, whereas that of PC is 62.5, 55.0 and 48.1 respectively. This is mainly because the social diffusion of PC and the PC-based Internet is substantially smaller than that of mobile phones in Japan. While the use rate of mobile phones reaches 71.9%, that of PCs is 56.7%, which is lower than even that of mobile internet.
The social distribution pattern for PC and mobile communications as described above motivated the Japanese government to make no distinction between the PC-based Internet and the mobile internet when taking statistics. Thus, the Japanese government’s analysis is not a multivariate one of PC-based Internet use, but a broader analysis of the internet via any access measures. However, given the data on PC use, the use of the PC-based Internet can be expected to be highly stratified by many different sociodemographic factors. In addition, the analysis of broadband internet use is limited to PC-based Internet users, which makes the figures for broadband use rate much higher than the analysis based on total respondents, since the PC-based Internet is used by only half of the respondents. In terms of the digital divide, however, it is far more important to analyze the distribution of broadband and FTTH among total respondents. Furthermore, the MIC’s analysis does not include education or occupation, which is generally a significant independent factor in the social diffusion of ICTs.
Therefore, the data from the JFK Survey is again examined to determine social distribution by the sociodemographic factors of sex, age, household income, education and occupation for the use of the following technologies: PCs, Internet via any access means, PC-based Internet, broadband and FTTH. In addition, the number of broadband users among PC-based Internet users is analyzed for the same factors. Since the JFK Survey was conducted exclusively in the 23 special wards of the city of Tokyo, the factor of city size is irrelevant.
Frequency distribution of these six variables and the results of their nominal logistic regression analysis are put in Table 10 and Table 11 respectively.
Table 10 Frequency distribution of PC, Internet, broadband use among the Japanese by sociodemographic factors based on JFK Survey
JFK Survey, Tokyo |
PC use |
Internet use (via any access means) |
PC-based Internet |
Broadband (among PC-based Internet users) |
BB (among total respon-
dents) |
FTTH (among total respon-
dents) |
total |
75.4 |
78.9 |
71.0 |
69.0 |
49.0 |
12.3 |
sex |
male |
79.1 |
82.6 |
76.5 |
74.4 |
57.0 |
14.8 |
female |
71.6 |
75.1 |
65.3 |
62.6 |
40.9 |
9.8 |
age |
20s |
91.1 |
95.6 |
86.7 |
73.1 |
63.3 |
13.3 |
30s |
83.1 |
94.4 |
79.8 |
72.7 |
58.1 |
14.5 |
40s |
81.4 |
85.6 |
78.4 |
61.8 |
48.5 |
11.3 |
50s |
67.1 |
64.6 |
61.0 |
72.0 |
43.9 |
13.4 |
60 and over |
38.7 |
32.3 |
32.3 |
55.0 |
17.7 |
6.5 |
annual household income |
under \2 million ($17,241) |
48.2 |
66.7 |
48.2 |
69.2 |
33.3 |
3.7 |
\2 to 4 million ($34,483) |
62.0 |
65.8 |
55.7 |
56.8 |
31.7 |
5.1 |
\4 to 6 million ($51,724) |
72.2 |
78.4 |
68.0 |
69.7 |
47.4 |
10.3 |
\6 to 8 million ($68,966) |
75.7 |
81.1 |
74.3 |
69.1 |
51.4 |
12.2 |
\8 to 10 million ($86,207) |
86.3 |
84.3 |
76.5 |
64.1 |
49.0 |
17.7 |
\10 to 12 million ($103,448) |
97.1 |
94.1 |
94.1 |
71.9 |
67.7 |
20.6 |
\12 to 14 million ($120,690) |
72.2 |
77.8 |
72.2 |
92.3 |
66.7 |
11.1 |
\14 million and over |
91.7 |
87.5 |
87.5 |
76.2 |
66.7 |
33.3 |
education |
elementary or junior high school |
31.6 |
42.1 |
26.3 |
60.0 |
15.8 |
10.5 |
high school |
58.8 |
64.2 |
52.7 |
61.5 |
32.4 |
8.8 |
junior college or technical college |
78.4 |
86.4 |
72.7 |
62.5 |
45.5 |
10.2 |
college and graduate |
92.1 |
90.0 |
89.4 |
73.8 |
66.0 |
16.0 |
occupation |
full-time worker |
83.6 |
84.7 |
79.1 |
68.4 |
54.1 |
13.8 |
part-time worker |
60.3 |
71.8 |
53.9 |
59.5 |
32.1 |
11.5 |
student |
100.0 |
100.0 |
100.0 |
82.4 |
82.4 |
5.9 |
housewife |
63.5 |
68.3 |
60.3 |
78.9 |
47.6 |
9.5 |
no occupation |
50.0 |
53\.6 |
46.4 |
61.5 |
28.6 |
7.1 |
Data: JFK Survey
Table 11 Multivariate statistical estimation of the relative weight of sociodemographic factors on PC, Internet, broadband use among the Japanese based on JFK Survey
|
|
sex |
age groups |
annual household income |
education |
occupation |
PC use |
ChiSq |
1.262 |
33.965 |
31.607 |
24.110 |
11.345 |
Prob>ChiSq |
0.2612 |
<.0001 |
0.0001 |
<.0001 |
0.045 |
PC-based Internet use |
ChiSq |
0.005 |
37.838 |
25.277 |
26.644 |
8.358 |
Prob>ChiSq |
0.9418 |
<.0001 |
0.0014 |
<.0001 |
0.1376 |
internet (via any measures) |
ChiSq |
0.960 |
86.097 |
16.399 |
11.576 |
1.698 |
Prob>ChiSq |
0.3273 |
<.0001 |
0.0370 |
0.0090 |
0.8892 |
BB access (among PC-based Internet users) |
ChiSq |
6.574 |
6.692 |
6.573 |
7.954 |
7.988 |
Prob>ChiSq |
0.0103 |
0.1531 |
0.5833 |
0.047 |
0.1569 |
BB (among total respondents) |
ChiSq |
5.679 |
31.761 |
20.166 |
19.294 |
6.879 |
Prob>ChiSq |
0.0172 |
<.0001 |
0.0097 |
0.0002 |
0.2298 |
FTTH (among total respondents) |
ChiSq |
3.186 |
4.975 |
18.766 |
2.317 |
10.188 |
Prob>ChiSq |
0.0743 |
0.2898 |
0.0162 |
0.5093 |
0.0701 |
Data: JFK Survey
The results show that age is again the strongest influential factor on PC use, Internet use, PC-based Internet and broadband among total respondents. Education and household income are almost as strong an influential factor on these uses as age. As expected, frequency distribution for PC-based internet use and broadband use among total respondents is stratified more broadly by age, household income, and education than internet use via any access means and mobile internet use. In addition, as for the use of FTTH, annual household income is the only factor statistically significant among the five sociodemographic factors, with a 0.05 significance level.
The phenomenon of mobile phones, PCs, PC-based Internet, and broadband Internet being distributed unequally by socio-demographic factors is true of not only Japan, but most industrialized societies as well. Eurostat conducted the “Community Survey on ICT Usage in Households and by Individuals” in 25 EU member states and candidates in the second quarter of 2004. Demunter (2005) gives some findings regarding the use of PCs and the Internet based on their frequency distribution depending on age, occupation, educational level, and the degree of urbanization of the area.
He points out that “(t)he digital divide is mainly a matter of age and education, whereas the gender gap is small” (Demunter 2005: 1). Since I compared the G7 countries, Finland, and Korea where possible, I provide the values for Germany, Italy, United Kingdom, Finland, and the average of the values from all 25 EU states in Table 12.
Although PCs and the Internet are used by more males than females to some extent except in Finland, the gender gap is no longer so prominent as it was in the early stages of ICT diffusion. According to the statistics, other socio-demographic factors appear to be much more significant. In particular, PC and Internet use decreases gradually with increasing age and falls dramatically among those aged 55 and over. Within the average data for the 25 EU states, the proportion of Internet users among those aged 16 to 24 was three times higher than that of those aged 55 to 74.
Table 12 Frequency distribution of PC and Internet use in selected EU countries by sociodemographic factors in 2004
|
computer use |
Internet use |
EU25 |
Ger-
many |
Italy |
UK |
Fin-
land |
EU25 |
Ger-
many |
Italy |
UK |
Fin-
land |
All individuals |
55 |
70 |
39 |
69 |
75 |
47 |
61 |
31 |
63 |
70 |
Sex |
Male |
58 |
73 |
45 |
72 |
75 |
51 |
65 |
37 |
67 |
70 |
Female |
51 |
66 |
34 |
66 |
75 |
43 |
57 |
26 |
59 |
71 |
Age |
Aged 16 to 24 |
82 |
96 |
69 |
85 |
97 |
75 |
92 |
58 |
83 |
96 |
Aged 25 to 34 |
69 |
89 |
55 |
85 |
94 |
62 |
85 |
46 |
80 |
93 |
Aged 35 to 44 |
63 |
83 |
47 |
78 |
89 |
54 |
74 |
37 |
69 |
85 |
Aged 45 to 54 |
52 |
74 |
38 |
72 |
75 |
43 |
61 |
28 |
63 |
68 |
Aged 55 to 64 |
34 |
48 |
18 |
52 |
62 |
27 |
36 |
12 |
42 |
53 |
Aged 65 to 74 |
15 |
25 |
4 |
27 |
19 |
11 |
17 |
3 |
23 |
12 |
education |
Lower educated |
31 |
59 |
18 |
36 |
58 |
25 |
51 |
13 |
28 |
54 |
Middle educated |
62 |
70 |
63 |
76 |
78 |
52 |
61 |
51 |
69 |
71 |
Higher educated |
84 |
84 |
79 |
91 |
92 |
77 |
76 |
71 |
87 |
89 |
occupation |
Employees |
70 |
84 |
54 |
81 |
87 |
60 |
74 |
42 |
74 |
82 |
Unemployed |
47 |
66 |
38 |
63 |
68 |
40 |
57 |
29 |
51 |
62 |
Students |
92 |
98 |
84 |
93 |
97 |
85 |
94 |
74 |
94 |
97 |
Retired |
18 |
31 |
9 |
28 |
26 |
13 |
23 |
6 |
24 |
20 |
area |
Densely-populated areas |
58 |
71 |
43 |
67 |
79 |
51 |
62 |
35 |
61 |
75 |
Intermediate areas |
55 |
69 |
37 |
75 |
72 |
48 |
61 |
29 |
68 |
65 |
Thinly-populated areas |
45 |
68 |
33 |
68 |
64 |
38 |
60 |
25 |
60 |
60 |
Source: Data extracted from Demunter 2005, Table 1, p.2.
Education is as important a factor as age. Within all Member States and Candidate States for which data is available, there is a higher level of Internet use among the higher educated. While more than 80 % of higher-educated EU citizens use PCs and three quarters of this population group use the internet, people with a lower level of education are only about one third as likely to use any of the two. Only 25 % of those who had not completed secondary school used the Internet, compared with 52 % of those who had attained a secondary school diploma.
Eighty five percent of students (aged 16 or higher in school or university) used the Internet during the first quarter of 2004, as did 60% of employees, 40% of the unemployed and 13% of the retired. In nearly all Member States, a lower proportion of the unemployed than employees used the Internet. In all Member States, the lowest rate of Internet use was observed for the retired. With respect to the degree of urbanization, computer and internet usage are still about 12 % lower in thinly-populated rural areas than in other areas within the EU.
Unfortunately, no analysis of the relative weight of these factors is made, nor is any data provided on their use by household income. And yet, the distribution of the use of PCs and the Internet by these socio-demographic factors in the EU is probably similar to that in Japan.
As for the United States, the Pew Internet & American Life Project provided three "thermometers" of digital access: internet, cell phone, and home broadband connections based on their survey conducted in April and May 2008. They provide their frequency distribution depending on sex, age, area, income, education and ethnicity (Table 13).
Sex makes no difference on Internet use or mobile phone ownership. The rate of broadband use among males is higher than that among females; it is not so significant as the other factors. The digital divide cuts along the same lines as in the EU and Japan: age and educational level. Americans aged 65 and older and those with less education lag behind in their adoption of the Internet and mobile phones. Similarly to Japan, annual household income seems as significant as age and education level.
One thing noticeable in the American data is the difference by area. The Internet, mobile phones, and broadband are all more widely used in suburban areas than urban areas in the United States. In most European countries and Japan, however the penetration of these technologies is higher among those living in urban areas than among those living in suburban areas. The United Kingdom is similar to the United States in this respect, which suggests inner city problems with both societies. Another noticeable difference in the American data is a digital divide along ethnic lines, which is relevant to many European societies but far less relevant to Japanese society.
Table 13 Rate of the Internet use, owing a cell phone and broadband use among the Americans
|
Internet use |
owning a cell phone |
using a high-speed connection (e.g., DSL, cable, wireless) when accessing the internet from home |
total |
73 |
78 |
55 |
sex |
male |
73 |
78 |
57 |
female |
73 |
78 |
52 |
age |
18-29 yrs old |
90 |
86 |
69 |
30-49 yrs old |
85 |
87 |
68 |
50-64 yrs old |
70 |
77 |
49 |
65+ yrs old |
35 |
55 |
19 |
area |
suburban |
77 |
81 |
59 |
urban |
74 |
75 |
56 |
rural |
63 |
73 |
38 |
income |
>$75K |
95 |
92 |
82 |
<$30K |
53 |
63 |
31 |
education |
college graduates |
91 |
89 |
78 |
high school graduates |
63 |
73 |
39 |
no high school degree |
44 |
59 |
29 |
ethnicity |
English-speaking Hispanics |
80 |
80 |
55 |
Whites |
75 |
79 |
56 |
Blacks |
59 |
73 |
41 |
Spanish-dominant Hispanics |
32 a |
|
|
Source: Date from Susannah Fox and Jessica Vitak “Degrees of Access (May 2008 data)” July 9, 2008. http://www.pewinternet.org/Presentations/2008/Degrees-of-Access-(May-2008-data).aspx, accessed January 10, 2010.
a. The statistics comes from the Pew Internet Project’s Latinos Online data, collected June-October, 2006.
Social distribution of mobile phones, PCs, the Internet, and broadband by socio-demographic factors in Korea is also similar to that in Japan, the EU and the United States. Since the Korean survey in JFK Survey was conducted in the same manner as its Japanese counterpart, the distribution of ICT use by sociodemographic factors is calculated (Table 14) and nominal logistic regression analysis is applied (Table 15).
Compared with the Japanese results (Table 8 and 11), though mobile internet use is far less widespread in Korea than in Japan, age is the only factor statistically significant, with a 0.05 significance level among the five socio-demographic factors both in Korea and in Japan. As in Japan, mobile phones are more widely used than PCs and PC-based Internet in Korea. In addition, the distribution of mobile phone use is much less stratified by age, income, and occupation in Korea. The use rate is close to 80% even among those aged 60 and over, and among those without a high school degree, (the lowest value for mobile phone use) it is over 70%.
Table 14 Frequency distribution of PC, Internet, broadband use among the Koreans by sociodemographic factors based on JFK Survey
|
mobile phone |
mobile internet |
PC use |
PC-based Internet |
Broadband (among PC-based Internet users) |
BB (among total respon-
dents) |
FTTH (among total respon-
dents) |
total |
92.8 |
9.7 |
70.4 |
68.8 |
86.6 |
59.6 |
8.0 |
sex |
male |
95.6 |
9.8 |
74.5 |
72.1 |
85.6 |
61.8 |
9.2 |
female |
90.0 |
9.6 |
66.3 |
65.6 |
87.8 |
57.5 |
6.9 |
age |
20s |
97.9 |
23.0 |
98.3 |
98.3 |
87.2 |
85.8 |
12.1 |
30s |
95.7 |
12.2 |
87.1 |
85.9 |
90.4 |
77.7 |
11.8 |
40s |
95.6 |
4.4 |
68.9 |
68.9 |
82.6 |
56.9 |
5.3 |
50s |
90.6 |
1.3 |
50.3 |
46.3 |
79.7 |
36.9 |
5.4 |
60 and over |
77.2 |
0.0 |
17.9 |
13.1 |
94.7 |
12.4 |
1.4 |
annual household income |
under 10 M wona |
90.6 |
6.3 |
46.9 |
43.8 |
85.7 |
37.5 |
6.3 |
10 to 20 M won |
88.0 |
4.0 |
53.0 |
52.0 |
84.6 |
44.0 |
7.0 |
20 to 30 M won |
89.8 |
7.6 |
61.8 |
57.8 |
86.9 |
50.2 |
8.4 |
30 to 40 M won |
92.9 |
8.3 |
72.1 |
71.7 |
84.1 |
60.2 |
7.5 |
40 to 50 M won |
97.5 |
13.1 |
79.4 |
75.0 |
89.2 |
66.9 |
6.9 |
50 to 60 M won |
99.1 |
13.0 |
85.2 |
87.0 |
90.0 |
78.3 |
11.3 |
60 to 80 M won |
96.3 |
13.0 |
87.0 |
88.9 |
77.1 |
68.5 |
9.3 |
80 M won and over |
97.4 |
18.4 |
81.6 |
84.2 |
93.8 |
79.0 |
7.9 |
education |
elementary or junior high school |
71.1 |
0.0 |
10.7 |
9.1 |
90.9 |
8.3 |
1.7 |
high school |
92.6 |
10.3 |
64.7 |
62.6 |
85.2 |
53.3 |
7.7 |
junior college or technical college |
97.0 |
12.0 |
85.0 |
84.4 |
87.9 |
74.3 |
10.8 |
college and graduate |
98.8 |
11.7 |
91.6 |
90.1 |
87.0 |
78.4 |
9.6 |
occupation |
full-time worker |
94.8 |
10.1 |
74.0 |
71.3 |
80.8 |
57.7 |
7.3 |
part-time worker |
95.5 |
10.5 |
65.7 |
65.7 |
88.6 |
58.2 |
7.5 |
student |
98.7 |
19.7 |
98.7 |
98.7 |
96.0 |
94.7 |
13.2 |
housewife |
88.7 |
5.4 |
60.6 |
60.6 |
96.7 |
58.6 |
8.9 |
no occupation |
79.7 |
6.8 |
43.2 |
43.2 |
100.0 |
43.2 |
6.8 |
Data: JFK Survey
a. The average exchange rate of Korean won to USD in 2005 was 1317.
Table 15 Multivariate statistical estimation of the relative weight of sociodemographic factors on PC, Internet, broadband use among the Koreans based on JFK Survey
|
sex |
age groups |
annual household income |
education |
occupation |
mobile phone use |
ChiSq |
4.323 |
5.037 |
15.546 |
21.137 |
3.396 |
Prob>ChiSq |
0.0376 |
0.2836 |
0.0494 |
0.0003 |
0.4938 |
mobile internet use |
ChiSq |
0.013 |
62.004 |
7.465 |
6.226 |
2.125 |
Prob>ChiSq |
0.9081 |
<.0001 |
0.4874 |
0.1829 |
0.7127 |
PC use |
ChiSq |
0.627 |
149.629 |
11.479 |
77.633 |
1.048 |
Prob>ChiSq |
0.4283 |
<.0001 |
0.176 |
<.0001 |
0.9024 |
PC-based Internet use |
ChiSq |
0.391 |
181.681 |
21.204 |
67.792 |
1.512 |
Prob>ChiSq |
0.532 |
<.0001 |
0.0066 |
<.0001 |
0.8246 |
BB access (among PC-based Internet users) |
ChiSq |
0.062 |
12.402 |
9.441 |
4.558 |
3.160 |
Prob>ChiSq |
0.8028 |
0.0146 |
0.3065 |
0.3357 |
0.5314 |
BB access (among total respondents) |
ChiSq |
3.554 |
125.760 |
16.489 |
47.422 |
23.970 |
Prob>ChiSq |
0.0594 |
<.0001 |
0.0359 |
<.0001 |
<.0001 |
FTTH (among total respondents) |
ChiSq |
4.429 |
16.173 |
2.230 |
3.718 |
6.046 |
Prob>ChiSq |
0.0353 |
0.0028 |
0.9731 |
0.4456 |
0.1958 |
Data: JFK Survey
When it comes to the use of PCs and various forms of PC-based Internet access, age is the strongest influential factor on all variables, i.e., the use of PC, PC-based Internet, broadband, FTTH, and broadband among PC-based Internet. Education is almost as strong an influential factor on these uses as age. Annual household income is not so strongly correlated with these uses as in Japan. Still, it is statistically significant, with a 0.05 significance level with PC-based Internet and broadband among total respondents.
When it comes to the digital divide within a society, almost the same pattern applies to Japan, Korea, the EU, and the United States. What is particular to Japanese society in terms of the development of the digital divide is that there are a considerable number of people who use the mobile internet without using the PC-based Internet. As discussed, the use of the mobile internet is not so influenced by sociodemographic factors (except for age), as the PC-based Internet. Furthermore, as mentioned previously, there are a considerable number of people who use the mobile internet without using the PC-based Internet. Therefore, the use of the internet via any access measures is also less strongly influenced by sociodemographic factors in Japan.
What kind of people, then, is prone to use the mobile internet without using the PC-based Internet? In order to answer this question, we divide the respondents into four groups: 1) those who use both the PC-based Internet and mobile internet, 2) those who use only the PC-based Internet, 3) those who use only the mobile internet, and 4) those who use neither. Frequency distribution for these four groups by sociodemographic factors based on the data of the JFK Survey and that of the MIC Communications Usage Trend Survey in 2005 is summarized in Table 16.
Table 16 Frequency distribution for four typologies of PC-based Internet use and mobile internet use by sociodemographic factors among the Japanese
|
JKF Survey |
The MIC Communications Usage Trend Survey in 2005 |
Both PCneta and Mnetb |
PCnet only |
Mnet only |
Neither of them |
Both PCnet and Mnet |
PCnet only |
Mnet only |
Neither of them |
Total |
52.8 |
18.2 |
7.9 |
21.1 |
45.2 |
10.4 |
10.4 |
30.9 |
sex |
male |
54.8 |
21.7 |
6.1 |
17.4 |
49.8 |
12.3 |
8.4 |
25.7 |
female |
50.7 |
14.7 |
9.8 |
24.9 |
40.9 |
8.7 |
12.4 |
35.8 |
age |
6 to 12 yrs old |
|
17.7 |
36.1 |
3.5 |
39.5 |
13 to 19 yrs old |
65.1 |
14.2 |
9.2 |
8.7 |
20s |
76.7 |
10.0 |
8.9 |
4.4 |
71.3 |
6.6 |
10.2 |
7.7 |
30s |
63.7 |
16.1 |
14.5 |
5.7 |
68.2 |
6.2 |
11.9 |
10.0 |
40s |
58.8 |
19.6 |
7.2 |
14.4 |
62.2 |
10.2 |
13.9 |
11.8 |
50s |
32.9 |
28.1 |
3.7 |
35.4 |
41.8 |
11.7 |
13.8 |
29.8 |
60 and over |
12.9 |
19.4 |
0.0 |
67.7 |
9.9 |
5.7 |
7.9 |
73.5 |
city size |
metropolitan areas |
|
50.3 |
11.7 |
9.7 |
25.2 |
cities |
43.7 |
10.0 |
10.7 |
32.6 |
towns and villages |
38.1 |
8.8 |
11.1 |
39.2 |
annual household income |
under \2 million ($17,241) |
33.3 |
14.8 |
18.5 |
33.3 |
24.6 |
9.0 |
15.1 |
47.8 |
\2 to 4 million ($34,483) |
46.8 |
8.9 |
10.1 |
34.2 |
31.6 |
7.4 |
13.8 |
45.3 |
\4 to 6 million ($51,724) |
45.4 |
22.7 |
10.3 |
21.7 |
42.3 |
10.0 |
12.5 |
31.7 |
\6 to 8 million ($68,966) |
55.4 |
18.9 |
6.8 |
18.9 |
47.6 |
11.9 |
10.0 |
27.1 |
\8 to 10 million ($86,207) |
56.9 |
19.6 |
7.8 |
15.7 |
51.4 |
11.6 |
9.3 |
25.0 |
\10 to 12 million ($103,448) |
73.5 |
20.6 |
0.0 |
5.9 |
57.6 |
11.4 |
6.0 |
21.6 |
\12 to 14 million ($120,690) |
61.1 |
11.1 |
5.6 |
22.2 |
|
\14 million and over |
62.5 |
25.0 |
0.0 |
12.5 |
education |
elementary or junior high school |
21.1 |
5.3 |
15.8 |
57.9 |
|
high school |
37.8 |
14.9 |
11.5 |
35.8 |
junior college or technical college |
55.7 |
17.1 |
13.6 |
13.6 |
college and graduate |
65.5 |
22.5 |
2.0 |
10.0 |
occupation |
full-time worker |
60.1 |
19.0 |
5.6 |
15.3 |
|
part-time worker |
39.7 |
14.1 |
18.0 |
28.2 |
student |
82.4 |
17.7 |
0.0 |
0.0 |
housewife |
42.9 |
17.5 |
7.9 |
31.8 |
no occupation |
21.4 |
25.0 |
7.1 |
46.4 |
a. PCnet stands for PC-based Internet.
b. Mnet stands for mobile internet.
First of all, I would like to underline the fact that around one out of ten people in Japan use the mobile internet only. It is almost impossible to find such a pattern of distribution in other industrialized countries. For example, examining the Korean data of the JFK Survey, the frequency distribution of the four groups is 9.6%, 59.2%, 0.1% and 31.1%. That is, the majority use the PC-based Internet only and almost none uses the mobile internet exclusively in Korea. In fact, only one out of the total 1,013 respondents belonged to the third group. When comparing between Japanese, Koreans and Finns in their twenties of the JFK Survey, we find those who use mobile internet only among the Japanese alone (Table 17).
Table 17 PC-based Internet use and mobile internet use in JFK Survey
|
Both PC-based Internet and mobile internet |
PC Internet only |
Mobile Internet only |
none |
Japanese Females |
73.2 |
7.3 |
14.6 |
4.9 |
Japanese Males |
79.6 |
12.2 |
4.1 |
4.1 |
Finnish Females |
17.5 |
82.0 |
|
0.6 |
Finnish Males |
32.4 |
65.7 |
|
1.9 |
Korean Females |
22.0 |
76.3 |
|
1.7 |
Korean Males |
24.0 |
74.4 |
|
1.7 |
Data: JFK Survey
These four groups among the Japanese are distinctive by sociodemographic factors. What we find of significance here is the contrast between the first group and the third group. The higher annual household income is, the larger the frequency of the first group is, but the smaller the frequency of the third group is. The same applies to education. The higher educational attainment is, the larger the first group is but the smaller the third group is. In both respects, annual household income and educational attainment, the third group is similar to the fourth group, i.e., those who fall through the net. As for occupation, part-time workers are far more likely to belong to the third and fourth groups than full-time workers and students. Thus, it can be said that while the first group is advantaged socio-economically, the third group and the fourth group are disadvantaged.
Given such differences among the four groups, it is of vital importance as to whether the PC-based Internet is significantly different from the mobile internet in terms of social concerns for the digital divide and if so, how and in what respect. This question is becoming increasingly important because the number of those who use the mobile internet exclusively seems to have grown in recent years. According to a survey conducted in November 2008, the frequency distribution of the four groups is 44.3%, 5.0%, 23.2% and 27.5% respectively. Thus the share of those who use only the mobile internet grew considerably from 2005 to 2008.
As mentioned above, there are two opposing positions. On the one hand, one might argue that there is no substantial difference between those who use PC based or both types of internet, and those who use only the mobile internet. According to its treatment of the statistics, the MIC is likely to hold this position. If so, the mobile internet would be considered to provide an alternative access measure to advanced communications networks for those who do not have or cannot afford access to a PC or PC-based Internet. In this sense, the mobile internet makes a substantial contribution to ameliorating the digital divide.
On the other hand, it could be claimed that the use and function of the mobile internet is significantly different from that of the PC-based Internet. Of course, the mobile internet is a powerful tool for communications and an innovative conduit for information circulation; it even leads to the development of a “mobile culture.” However, the mobile internet can never completely substitute for the PC-based Internet. What we need is both forms of access. If society does not recognize the substantial differences between these two technologies and leaves those who use the mobile internet as they are, the issue of the digital divide will remain far from solved. Instead, the real issue of the digital divide will continue to be obscured by the high penetration rate of the mobile “internet,” which poses the dangerous risk of a false sense of security and subsequent social complacency.
After my longitudinal involvement in research on the diffusion of ICTs and their use in Japan as well as in other societies, I believe that unfortunately, the latter is the case in this country. One of the most serious social consequences of neglecting this problem is an uneven distribution of “information literacy.”
The term and concept “information literacy” has drawn much discussion and research since the 1970s. Despite its importance as a concept, it is difficult to define, surrounded by extensive debate. Thus, in Japan, scholars of information society studies have attempted to develop an index to measure information literacy operationally, so that we might grasp its social distribution according to various socio-demographic factors and keep track of its change over time (Shozugawa 1996). Working with some of these individuals, I developed the IHC index, or the Information Handling Competency index.
The index is calculated based on five questions. The first four are based on a sequential model for people’s behavior when dealing with information: “searching for information,” “gathering information,” “selecting information” and “conveying information.” The fifth question concerns “collaborative skill” because collaboration is vital for communication via information networks. Below is a list of the five questions.
1. Searching for information: I like searching for information that interests me by myself, even though it costs considerable effort to find it.
2. Gathering information: I have my own method for gathering information.
3. Selecting information: I can sort the information I need out of a great amount of information.
4. Conveying information: I can convey information appropriate to the situation when I make conversation with others or while doing business.
5. Collaborative skill: I like to collaborate with someone else when creating something new.
The IHC index is calculated as follows. The response “agree” to each question is scored as two points, the response “somewhat agree” as one point, and the responses “somewhat disagree” and “disagree” as zero. The IHC index is defined as the simple addition of each score of five questions, ranging from 0 to 10.
The IHC index has been tested in several surveys in which I have been involved, yielding consistent and striking results: the IHC index differs significantly according to whether or not the respondent uses the PC-based Internet. In the previous chapter, I pointed out that there are a considerable number of people who use the mobile internet without using PC-based Internet in Japan.
Within these surveys, the IHC index has measured significantly higher among PC-based Internet users than among those who use mobile internet alone or among those who use neither types of network, even after controlling for other factors such as age, sex, education, and annual income (Kimura 2000, 2002, 2004). Table 18 summarizes the IHC index of the four groups (those who use both PC-based Internet and mobile internet, those who use PC-based Internet alone, those who use mobile internet alone and those who use neither) in six different surveys and the distribution of the four groups there. The six surveys vary considerably in terms of their date, their area, the age range of their samples and the distribution of the four groups. However, the IHC index value for each of the four groups remains remarkably stable.
The result shows that although the Japanese government counts those who use mobile internet only as “Internet users,” the mobile internet is far from being the same as the PC-based Internet in terms of the degree of information literacy required. As shown in Table 16 and 17, there is a considerable number of people who do not use the PC-based Internet even among those in their 20s and 30s, who tend to be low-income, less educated, and without occupation. Because the PC-based Internet provides people with a high degree of convenience, efficiency, and opportunity, those who do not have access to the PC-based Internet are all the more disadvantaged. The fact that the mobile communications network and mobile phones have advanced to the point of generating such a considerable number of people who use the mobile internet exclusively has only resulted in the serious social consequences caused by a digital divide particular to Japanese society.
Table 18 IHC index according to the use pattern of PC-based Internet and mobile internet in different surveys
|
RISS Survey in 2000 |
Seikei Survey in 2001 |
Panel survey in 2001 |
Panel survey 2003 |
JFK Survey in Japan |
JFK Survey in Korea |
IHC index |
Both PCneta and Mnetb |
5.65 |
4.93 |
5.02 |
5.02 |
5.18 |
4.62 |
PCnet only |
4.97 |
5.18 |
4.98 |
4.96 |
3.89 |
4.36 |
Mnet only |
3.83 |
3.18 |
3.83 |
3.76 |
3.36 |
2.00 |
Neither of them |
2.98 |
3.57 |
3.58 |
3.38 |
2.77 |
2.93 |
distribution of the four groups |
Both PCnet and Mnet |
3.0 |
70.5 |
21.9 |
29.1 |
52.7 |
9.6 |
PCnet only |
18.2 |
12.2 |
15.7 |
14.7 |
18.2 |
59.2 |
Mnet only |
3.0 |
12.2 |
17.7 |
17.2 |
7.9 |
0.1 |
Neither of them |
75.8 |
5.0 |
44.7 |
39.1 |
21.1 |
31.1 |
a. PCnet stands for PC-based Internet.
b. Mnet stands for mobile internet.
So far, we have discussed the access divide. However, as the Internet and mobile communications have been diffused broadly in society, more attention has been paid to the divide in use as well. When we turn to the divide in use, Japanese society is characterized as having a low level of online activities compared with other societies.
The World Internet Project investigates a variety of online activities and services, classifying them into five categories. The first is that of communication services, such as e-mail, instant messaging, and online telephone calls; the second is information services, such as news and travel information sites; the third is online activities for fun, such as online games and downloading and watching videos; the fourth is e-commerce and transaction services, such as online purchasing and online banking; and the last is online learning activities, such as distance learning.
Mikami (2009) conducted a principal component analysis of the social differences in such online activities and services in Japan and twelve other countries where WIP surveys were conducted in 2007, i.e., Australia, Canada, China, Colombia, Czech Republic, Hungary, Israel, Macao, New Zealand, Singapore, Sweden, United Kingdom, and the United States. 18 items of online activities and services are selected and their use rate for each society among Internet users is put into principal component analysis.
As a result, five components with over one point of eigenvalue are distinguished. The contribution rate of these components is 32.2%, 19.4%, 13.1%, 11.9% and 8.4% respectively. The first and second components explain over half of the variance. Mikami provides the loadings of 18 items and the scores of 13 countries in these two components (Table 19).
Those items belonging to online activities for fun (the third category), information services (the second category), and synchronous communications, instant messaging, and online chat (the first category) are of high loadings in the first component. E-mail and those items belonging to the fourth category, e-commerce and transaction services, are of high loadings in the second component. The factor scores for the countries in the first component show that online activities for fun, use of information services, and synchronous communications are vigorous in China, the Czech Republic, and Israel, and in low gear in Colombia, Japan, and Sweden; those in the second component suggest that transaction services are actively used in New Zealand, Australia, the Czech Republic, the United States, and Canada, and are not used much in Hungary, Colombia, Japan, and China. Japan is the most inactive in online activities or use along with Colombia.
Social surveys I have been involved in consistently prove that the Japanese use the most advanced and sophisticated ICT environment mostly for seeking information for consumption, for watching videos and downloading music for fun, for peeping into BBSs dominated by anonymity, and for making social communications with those already acquainted. The Internet is rarely used for social participation, political engagement, accessing to online government, online transactions, online learning, telemedicine, self-expression, or expanding social networks. Just as the use of voice phone call is remarkably limited, so are synchronous communications such as Skype, instant messaging and online chatting via PC-based Internet among the Japanese.
Table 19 Result of principal component analysis of the social differences in online activities and services comparing 13 countries based on WIP data
Factor scores of online activities and services |
Factor scores of countries |
1st component |
2nd component |
1st component |
2nd component |
Games |
0.904 |
Banking Services |
0.868 |
China |
5.008 |
New Zealand |
2.377 |
Music |
0.868 |
E-mail |
0.868 |
Czech |
3.545 |
Australia |
2.177 |
Health information |
0.860 |
Bills |
0.863 |
Israel |
2.617 |
Czech |
1.945 |
Sexual Content |
0.851 |
Purchasing |
0.587 |
Singapore |
0.442 |
US |
1.896 |
Videos |
0.819 |
Travel Reservations |
0.462 |
US |
-0.016 |
Canada |
1.473 |
Looking for jobs or work |
0.718 |
Travel information |
0.367 |
Hungary |
-0.109 |
UK |
0.146 |
Instant messaging |
0.656 |
Distance Learning |
0.318 |
Canada |
-0.761 |
Israel |
-0.241 |
Chat |
0.617 |
News |
0.295 |
UK |
-0.779 |
Singapore |
-0.471 |
Travel information |
0.604 |
Sexual Content |
0.238 |
Australia |
-0.789 |
Sweden |
-0.509 |
Travel Reservations |
0.357 |
Fact-finding |
0.224 |
New Zealand |
-1.124 |
Hungary |
-1.021 |
News |
0.314 |
Look up a Word |
0.121 |
Colombia |
-2.393 |
Colombia |
-2.427 |
Distance Learning |
0.257 |
Looking for Jobs or works |
0.086 |
Japan |
-2.808 |
Japan |
-2.432 |
Purchasing |
0.195 |
Games |
0.007 |
Sweden |
-2.833 |
China |
-2.914 |
Look up a Word |
0.050 |
Health information |
-0.023 |
|
Fact-finding |
0.036 |
Chat |
-0.058 |
Bills |
0.002 |
Instant messaging |
-0.150 |
E-mail |
-0.133 |
Videos |
-0.300 |
Banking Services |
-0.187 |
Music |
-0.327 |
Data: WIP survey
Such a low level of use has much to do with the strong apprehension towards cyberspace at large (see Kimura 2004). In the JFK Survey, we asked to what extent people worry about various risks associated with Internet use (Table 20). Each of the items below has a five-category scale, from very afraid to not afraid at all. As Table 20 indicates, young Japanese, especially females, worry greatly about such risks.
Table 20 Frequency distribution of the question items related to risks of Internet use, JFK Survey
In using the Internet, are you afraid that ... |
Japanese females |
Japanese males |
Finnish Females |
Finnish males |
Korean females |
Korean males |
(a) Others will come to know the web sites you have seen and what you write in e-mail messages? |
very afraid and afraid to some extent |
78.1 |
69.4 |
47.6 |
43.9 |
75.4 |
58.7 |
hard to say |
9.8 |
14.3 |
11.3 |
16.1 |
16.1 |
20.7 |
not so afraid and not afraid at all |
12.2 |
16.3 |
41.0 |
40.0 |
8.5 |
20.7 |
(b) Others can find out your name and address or credit card number, etc.? |
very afraid and afraid to some extent |
97.6 |
89.8 |
72.2 |
61.6 |
85.6 |
74.4 |
hard to say |
0.0 |
6.1 |
5.5 |
9.4 |
11.0 |
14.9 |
not so afraid and not afraid at all |
2.4 |
4.1 |
22.3 |
29.0 |
3.4 |
10.7 |
(c) You will get your PC infected with a computer virus? |
very afraid and afraid to some extent |
97.6 |
71.4 |
83.0 |
64.2 |
74.6 |
68.6 |
hard to say |
0.0 |
12.2 |
4.4 |
8.1 |
17.8 |
17.4 |
not so afraid and not afraid at all |
2.4 |
16.3 |
12.7 |
27.7 |
7.6 |
14.1 |
(d) The information on your computer and the contents of your hard drive will be stolen? |
very afraid and afraid to some extent |
95.1 |
71.4 |
49.4 |
40.6 |
71.2 |
62.0 |
hard to say |
2.4 |
16.3 |
13.5 |
14.8 |
18.6 |
19.8 |
not so afraid and not afraid at all |
2.4 |
12.2 |
37.1 |
44.5 |
10.2 |
18.2 |
(e) Your passwords necessary for various Internet uses will be stolen and abused? |
very afraid and afraid to some extent |
97.6 |
75.5 |
57.9 |
52.6 |
83.1 |
70.3 |
hard to say |
2.4 |
12.2 |
11.7 |
9.7 |
11.0 |
16.5 |
not so afraid and not afraid at all |
0.0 |
12.2 |
30.4 |
37.8 |
5.9 |
13.2 |
|
Unit: % |
Data: JFK Survey II
Along with such a cautious attitude toward the Internet as a social communication medium is a strong concern with online encounters and recognition of the Internet as an anonymous communication space. Korean people involved in cyberspace use their real names. Social networks in the real world and in cyberspace overlap and reinforce one another and cyberspace functions as a means of communication and social activity. As is widely known, the reigning service in Korean internet use is the “community site.” Community sites such as “daum” (www.daum.net) and “cyworld” (www.cyworld.com) have more than 10 million registered users and millions of communities. People are expected to disclose themselves and use their real names (see also Kim 2004).
In contrast, among the Japanese, cyberspace is dominated by the notion of “anonymity"; people tend to avoid being involved in social communication over the Internet. When they engage in social communications on BBSs, blogs or social networking sites, they prefer to be anonymous and they do not disclose much about themselves. Anonymous edits of articles in Wikipedia are telling (Table 21). An edit of an article in Wikipedia can be made either with the name of the registered account or anonymously; in the case of the latter, the IP address from which the anonymous edit is made is recorded and shown on the article and its edit history. I collected the data from Wikipedia sites of several different languages with substantial numbers of users. It turns out that Japanese Wikipedia is outstanding in that the number of anonymous edits is almost half of the total edits. This is even more astounding because even English Wikipedia, where the most heterogeneous population is likely to gather and vandalism reigns over a great number of contentious and divisive articles, only 30 % of total edits are anonymous. In other populous language versions, the rate of anonymous edits is from 10% to 20% mostly.
Table 21 The number of anonymous edits and its percentage of total edits in different language versions of Wikipedia
|
The number of anonymous edits, i.e., with IP address (thousand) |
The number of total edits (thousand) |
Percentage of anonymous edits |
Japanese |
6,831 |
14,517 |
47.1% |
English |
17,779 |
57,880 |
30.7% |
German |
7,664 |
27,083 |
28.3% |
French |
3,373 |
17,923 |
18.8% |
Dutch |
1,094 |
8,087 |
13.5% |
Spanish |
3,198 |
10,531 |
30.4% |
Finnish |
612 |
2,924 |
20.9% |
Russian |
679 |
4,965 |
13.7% |
Chinese |
664 |
4,221 |
15.7% |
Korean |
117 |
985 |
11.9% |
Note: the data of the English Wikipedia is as of October, 2006; the data of the others is as of February 1st, 2008.
8. Conclusion
Examining the “Falling through the Net” in Japan reveals that the world’s most advanced and cutting-edge ICT equipment and services have been developed and commercialized aggressively in this country. However, due to a huge amount of investment in R&D and commercial deployment of such services, prices and charges for ICT services tend to be rather high. Thus, a considerable number of people fall through the network; as far as basic technological diffusion, such as the overall Internet penetration rate or the mobile phone penetration rate, is concerned, Japan falls in the middle among the countries of OECD, Organisation for Economic Co-operation and Development.
In addition, the Japanese government has prioritized the development of an advanced ICT infrastructure, especially concerning the establishment of a trunk network of fiber optics, while neglecting the need to stimulate the social use of ICT applications and human resource development. Thus, while advanced ICT trunk networks are developed nationwide and cutting edge services such as FTTH and 3G mobile phones are available to those who can afford them, the development of information literacy and skills is considered a matter of those who need special care and the fact that many citizens with low-income or low educational attainment are likely to fall through the net is not regarded as a social issue to be addressed (Kimura 2004).
One of the most serious social concerns about such digital divide in Japanese society is the fact that a considerable number of people who use mobile internet without using PC-based Internet in Japan. Social surveys I have been involved in consistently prove that information literacy has been significantly higher among PC-based Internet users than among those use mobile internet alone or among those use neither of them, even controlling other factors such as age, sex, education and annual income.
When it comes to the digital divide in ICT use among the Japanese, compared with the Koreans, the Finns, and the Americans, we have found that even though advanced features of ICTs are commercialized and available for use in Japan, the actual usage of various features and functions is very low. Furthermore, what has struck us the most about the way ICT is used in Japan is the strong apprehension towards cyberspace and the recognition of the Internet as an anonymous communication space. In this sense, Japanese society falls on the excluded side of the digitally divided from an international perspective.
Carvin, A.
2000a Mind the Gap: The Digital Divide as the Civil Rights Issue of the New Millennium. MultiMedia Schools. 7(1):56-58. http://www.infotoday.com/MMSchools/jan00/
carvin.htm, accessed December 28, 2009.
2000b More Than Just Access: Fitting Literacy and Content into the Digital Divide Equation. Educause Review 35(6):38-47. http://www.educause.edu/pub/er/
erm00/articles006/erm0063.pdf, accessed on December 28, 2009.
Cooper, Mark
2004 Expanding the Digital Divide & Falling Behind on Broadband: Why a Telecommunications Policy of Neglect is not Benign. Consumer Federation of America. http://www.consumerfed.org/pdfs/digitaldivide.pdf, accessed December 28, 2009.
Davidson, E., and S. R. Cotton
2003 Connection Discrepancies: Unmasking Further Layers of the Digital Divide. First Monday (8)3. http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/
1039/960, accessed January 10, 2010.
Demunter, Christophe
2005 The Digital Divide in Europe. Statistics in Focus, Industry, Trade and Services, 38. Brussels: European Communities.
Dewan, S., and F. J. Riggins
2005 The Digital Divide: Current and Future Research Directions. Journal of Association for Information Systems 6(2):298-337.
DiMaggio, Paul, Eszter Hargittai, C. Celeste, and S. Shafer
2004 From Unequal Access to Differentiated Use: A Literature Review and Agenda for Research on Digital Inequality. In Social Inequality. Kathryn M. Neckerman, ed. Pp. 355-400. New York: Russell Sage Foundation.
DiMaggio, Paul, and Eszter Hargittai
2001 From the ‘Digital Divide’ to `Digital Inequality’: Studying Internet Use As Penetration Increases. Working Paper #15. Princeton: Center for Arts and Cultural Policy Studies, Princeton University. http://www.princeton.edu/~artspol/workpap/
WP15%20-%20DiMaggio%2BHargittai.pdf, accessed December 28, 2009
Hargittai, Eszter
2002 Second Level Digital Divide: Differences in People’s Online Skills. First Monday 7(4). http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/ fm/ article/view/942/864,
accessed January 10, 2010.
2003 Informed Web Surfing: The Social Context of User Sophistication. In Society Online: The Internet in Context. Philip N. Howard and Steve Jones, eds. Pp.257-274. Thousand Oaks, CA: Sage Publications.
Kim, Sangmi
2004 Online Communication Riyou to Shakai-teki Skill: Ni-kkan Daigakusei Chousa wo Chu-shin ni [Online communication use and social skill: Focusing on comparative research on the Japanese and Korean university students]. Journal of Socio-Information Studies 8(2):13-26.
Kimura, Tadamasa
2000 Dejitaru Devaido: Sekai System no Sotai teki Sai Hensei to Nihon Shakai no Sentaku [Digital Divide: Comprehensive Restructuring of World System and Decision to Be Made of Japanese Society]. Memoirs of Tokyo Metropolitan Institute of Technology 14:1-6.
2002 Dejitaru Devaido no Hikaku Shakai Bunka ron [Socio-Cultural Comparative Study of "Digital Divide"]. In C&C Promotion Foundation, ed. Pp. 25-101.
2004 Netto-waaku Riaritii: Posuto Koudo Shohi Shakai wo Yomi-toku [Network reality: Exploration of post advanced consumer society]. Tokyo: Iwanami Shoten.
Kling, R.
1999 Can the “Next-Generation Internet” Effectively Support “Ordinary Citizens”? The Information Society 15:57-63.
Loader, Brian D., and Leigh Keeble
2004 Challenging the Digital Divide? York: Joseph Rowntree Foundation.
MIC
2006 Heisei 17 nen Tsushin Riyo Doko Chosa Setai hen [Report on Communications Usage Trend Survey (CUTS) 2005, Household Data]. Tokyo: MIC.
Mikami, Shunji
2009 Kokusai Hikaku Data no Bunseki [International comparative analysis]. In 2008 nen Internet Riyou ni Kansuru Jittai Chosa [Research on the Internet use in 2008]. Ken Sakamura and Yoshiaki Hashimoto, eds. Pp. 102-129. Tokyo: University of Tokyo.
NTIA
1995 Falling Through the Net: A Survey of the ‘Have Nots’ in Rural and Urban America. Washington, D.C.: National Telecommunication Information Administration, United States Department of Commerce.
1998 Falling Through the Net II: New Data on the Digital Divide. Washington, D.C.: National Telecommunication Information Administration, United States Department of Commerce.
1999 Falling Through the Net: Defining the Digital Divide. Washington, D.C.: National Telecommunication Information Administration, United States Department of Commerce.
2000 Falling Through the Net: Toward Digital Inclusion. Washington, D.C.: National Telecommunication Information Administration, United States Department of Commerce.
Rice, R. E., and J. E. Katz
2003 Comparing Internet and Mobile Phone Usage: Digital Divides of Usage, Adoption, and Dropouts. Telecommunications Policy (27)8/9:597-623.
Shukunami, Tatsushiro
2006 Broadband Service ni okeru Dejitaru Devaido no Kaisho Seisaku ni tsuite [On the policy to overcome the digital divide of broadband service]. Media Communication (Institute for Media and Communications Research, Keio University) 56:143-156.
Tolbert, Caroline J., and Karen Mossberger
2006 New Inequality Frontier: Broadband Internet Access. Economic Policy Institute Working Paper No. 275. Washington, DC: Economic Policy Institute.
|