Page 231 - Information and American Democracy Technology in the Evolution of Political Power

P1: IBE/IRP/IQR/IRR
CY101-Bimber
August 13, 2002
CY101-05
0 521 80067 6
Political Individuals 12:12
The term “digital divide” has come into wide use to describe dif-
ferences at the individual level between those who are “on line” and
those who are not. (No comparable attention is paid to gaps at the level
of political organizations, perhaps in large part because few such gaps
exist.) Since the individual-level digital divide mirrors traditional socio-
economic gaps in the United States, many observers, including agencies
of the federal government, have rightly expressed concern that new in-
formation technology may tend to exacerbate inequality in American
society. 38 These concerns have driven a number of federal and state
policies aimed at bringing the Internet to poor areas and underprivi-
leged schools, as is also happening in Europe and other places around
the world. 39
In practice, several “divides” are associated with information technol-
ogy at the individual level. One involves simply access to the Internet.
Another involves gaps in the kinds of habits and skills that lead to regular
use of the technology, while yet another involves systematic differences
across nations in access and use of the technology. 40 Problems of social
inequality are no less complex in matters of technology than in other
areas, and it is beyond the scope of the present study to assess the nature
of digital divides in detail. Nonetheless, it is possible to provide a rough
assessment of their structure here. Comparing the last two columns in
Table 5.1 suggests that the most important differences between Inter-
net users and nonusers involve education and race. Since these variables
are themselves correlated, it is important to analyze matters of the digi-
tal divide using a multivariate approach. Table 5.2 presents the results of
logistic regression analysis predicting two variables: access to the Internet
41
and daily use of the Internet among adults with access. The indepen-
dent variables measure education, age, household income, sex, ethnic-
ity, race, whether the respondent is employed full-time, whether the
respondent is married, and whether the respondent is a full-time home-
maker. The factors predicting variation in access are education, income,
age, Latino ethnicity, and homemaker status. The model is reasonably
strong, indicating that demographic information is moderately useful

38
U.S. Department of Commerce, National Telecommunications and Information Ad-
ministration, Falling Through the Net: Defining the Digital Divide (Washington, D.C.:
Government Printing Office, 1999), http://www.ntia.doc.gov/ntiahome/fttn99.
39 40
Norris, Digital Divide. See ibid.
41
The most useful units for measuring frequency of Internet use have shrunk over time,
as citizens have made greater use of new information technology: Between 1996 and
1999, I measured frequency of Internet use using days as units of measure; beginning
in 2000, I changed to hours per day.
214