Abstract
The increasing spread of the Internet holds much potential for enhancing opportunities for people with disabilities. However, scarce evidence exists to suggest that people with disabilities are, in fact, participating in these new developments. Will the spread of information technologies (IT) increase equality by offering opportunities for people with disabilities? Or will a growing reliance on IT lead to more inequality by leaving behind certain portions of the population including people with disabilities? In this paper, the authors draw on nationally representative data regarding Americans' Internet uses to (1) identify the extent to which people with disabilities are embracing use of the Internet; (2) how their use of the Internet compares with the Internet uses of the rest of the population; (3) how having a disability relates to and interacts with other social statuses (e.g. socioeconomic status, age, gender) with regard to Internet use; and (4) what explains these trends. They draw on representative data collected by the Bureau of Labor Statistics and the Census of the United States to answer these questions. It is found that people with disabilities are less likely to live in households with computers, are less likely to use computers and are less likely to be online. However, once socioeconomic background is controlled for, it is found that people with hearing disabilities and those who have limited walking ability are not less likely to be Internet users. This research enables a deeper understanding of both the use of the Internet by people with disabilities and the spread of new IT more generally.
Acknowledgements
The authors would like to thank Jennifer Humensky for help with data management. Jeremy Freese and Peter Miller provided helpful suggestions. Hargittai acknowledges the support of the Northwestern University Communication Studies Department Research Fund.
Notes
1 A simple search on two popular discussion group aggregator sites (Yahoo! Groups and Google Groups) yields well over a thousand mailing lists devoted to such topics.
2 Among Latinos, the differences in Internet access rates are not significantly different between those with and without disabilities (Kaye Citation2000).
3 Ideally we would have even more refined data, but this is the extent to which the topic is covered in the data set. For example, there are no questions about the availability and use of assistive technologies.
4 We performed a log transformation on the income variable, because it makes sense to assume that the payoff of additional dollars diminishes as one moves up the scale. Moreover, this was important to meet the normal distribution requirement for logistic regression analysis.
5 In another model (not reported) with the same control variables, we included a variable for the number of disabilities the respondent reports (logged so as to give less weight to each additional disability). The variable was significant at the p < 0.000, indicating that the more disabilities an individual has, the more likely he or she is to face barriers to Internet access and use.