An analysis of the 2002 SIQSS time-diary data offers support for the displacement theory of time utilization. Time spent on the Internet is found to have a negative relationship with a number of daily activities, especially discretionary activities. Most notably, time spent on the Internet appears to come at the expense of time spent on social activities, hobbies, reading and TV viewing. Time spent on the Internet has a small, but less substantive, impact on time spent on work, childcare, housework and sleep. These predictions are robust to multivariate controls and proportional comparisons.

According to the "functional equivalence" argument that has been applied to the diffusion of earlier communication technologies, one should expect decreases in daily activities that perform the same functions as the Internet. An effective, comprehensive method for testing which activities seem most affected by the Internet is through 24-hour time-diary studies, in which all daily activity is recorded.

When one compares the time diaries of Internet users and nonusers in a combined 1998-2001 national telephone diary study, one finds consistent declines in TV use and sleep times among Internet users, but no consistent declines in reading or other activities. This finding suggests some ways that the Internet may affect some communication activities more than others. This lack of strong changes is a rather different pattern from what seems to have been the case for television, as predicted by the functional-equivalence hypothesis.

This study compared the time use of parent and child IT users with that of parent and child nonusers. The study took advantage of a year 2000 data collection that involved a national sample of 450 dual-income, middle-class families. All family members kept a complete, weekly account of all of their daily activities. The sample thus has the advantage of being restricted to a relatively homogeneous population group in terms of age, family circumstance and life stage.

The major difference found between parent Internet users and nonusers is in terms of the 5+ lower paid work hours of Internet users that offset their 3.5+ weekly hours of Internet use. Parent Internet users slept significantly less, but spent more time reading, radio/music listening and engaging in hobbies. Among children in these households aged 5-18, the major differences between Internet users and nonusers after multivariate adjustments are found in their slightly greater time attending social events and engaging in conversation. Otherwise, the diary figures of Internet users and nonusers, especially for children, are strikingly similar.

Internet use data were collected in the 1998 time-use survey of Statistics Canada, done as part of their General Social Survey. That survey collected single-day diary data from 10,749 persons aged 15 and older by telephone, during the entire calendar year, with a response rate of 78 percent. Previous Canadian diary results matched U.S. results rather closely.

It was found that respondents aged 18-64 who reported IT use in their day's diary reported less sleep than nonusers, but no less time watching TV, and more time reading, corresponding and doing hobbies; these differences largely held up after adjustment for several, demographic predictors. In contrast, when the analysis focused on longer-term users, users watched significantly less TV than nonusers, but sleeping times were the same.

The impact of the Internet is put into the context of long-term, time-use trends in the United Kingdom, taking advantage of almost 40 years of time-diary studies. The trends in time use that emerge challenge several popular beliefs about how society has been changing in the wake of new technology and other social changes since WWII. The main focus of the analysis is on a unique set of panel data in which respondents in nearly 1000 households completed full-week diaries in 1999 and 2000, making it possible to distinguish changes in three groups: prior Internet users, nonusers at both time points and new Internet users. No notable changes in media use or other activities were found in any of the three groups across the span of the study, indicating little support for the hypothesis that Internet users had significantly altered their styles of life.

Differences in time use between Internet users and nonusers in the Netherlands are found mainly in the productive activities of paid work, childcare and household chores. Internet users spend two hours less per week on paid work and almost two hours less on childcare and domestic tasks. On the other hand, because more Internet users are students, they spend an hour more on education than nonusers. Thus, Internet users have approximately three hours more free time than nonusers. However, hardly any differences in specific free-time activities can be identified, except that users spend more time on reading books and visiting cultural institutions and nonusers spend less time attending to children and helping adults.

Data from the 2000 Time Use Survey (TUS; n=1813) are used to analyze the composition of media repertoires of the Dutch population, with particular reference to the Internet. Based on an entire media supply of 19 items, factor analysis identified eight factors underlying the repertoires. Multiple regression analysis confirmed that the eight repertoires identify audiences that are clearly distinct in terms of status, gender, age, labor market position, as well as religion and political interest.

Two separate groups of Internet and PC users were discerned: 1) Users of the Internet for serious information, who are young, over 70% male, with slightly above-average schooling level and also above-average cultural and economic occupational status; the group works long hours but spends little time on household chores. They combine using the Internet with using a PC (offline) and reading hobby magazines. 2) Users of the Internet for purposes other than serious information; who are more often female (40%), and also more highly educated and of a higher cultural job status.

The year 2000 General Social Survey (GSS) included old questions on the extent of TV viewing and newspaper reading that provide bivariate support for the conclusion that Internet users watch less TV than nonusers, but that this difference is largely explained by demographic differences between the two groups. On the other hand, Internet users are more likely to read newspapers, and significantly so after multivariate adjustment for demographic predictors-but these differences are not monotonic with extent of usage.

The 2000 GSS also asked new questions on the use of the Internet vs. other media for information on health and politics. Analysis of these comparative media questions also provides little evidence of reduced usage of traditional media by those who use the Internet for health or political content, particularly for heaviest Internet users for those purposes. There was more evidence to support the "Newtonian" model of increased media use among Internet users than evidence of any displacement effect.

In this secondary analysis of the UCLA year 2000 data, an MCA multivariate approach is used to test the relation between Internet use, mass media and other activities, using both behavioral and attitudinal data in the UCLA surveys. These analyses indicate little consistent evidence of significantly decreased media or other behavior among heavier Internet users, after age, education, income, race, gender and marital status are taken into account. Internet users did report 5 hours less TV viewing, but only 1.5 hours less after MCA adjustment, and the differences were not monotonic with amount of Internet use.

On the contrary, there is significant evidence of more reading of books, video game playing andb music listening among Internet users, even after adjustment for these same demographic factors. Radio listening and reading newspapers/magazines were also high among heavier Internet users, but the relationships were neither monotonic nor significant. Subjectively, Internet users did rate less importance than nonusers for TV, radio, organizations and colleagues as sources of information-but not for entertainment.

Revolutionary information technology has the potential to affect usage of other media and other ways of spending time. Using data from a national random sample of 4113 adults, it is found that the more respondents report using the Internet, the more they report reductions in their time watching TV, reading newspapers, shopping in stores and driving in traffic-and the more time they spend working both at home and the office. The changes are greater for those using the Internet progressively more.

The Internet is often characterized as a "disruptive technology," as recently argued by the music industry against Napster and by Sony against PC emulations of its PlayStation. Three questions are raised: 1) How much does the Internet replace traditional media? 2) As another information channel, does the Internet supplement traditional media? 3) How much do Internet users navigate the Web in order to download relevant information goods (news, music, movies) as part of their consumption and purchase strategies?

Data from the Pew Internet and American Life Project are used to test two competing hypotheses: 1) Easily copied and distributed digital content encourages consumers to exploit the "information wants to be free" character of the Internet, thereby hampering the growth of the market for information goods and services, and 2) Information consumers practice a sophisticated arbitrage process across different media, by weighing the value of online and offline information and thereby reinforcing active participation in the electronic marketplace.

The efforts of the music industry to obstruct the downloading of MP3s may represent a serious misunderstanding of consumer arbitrage practices, much as book publishers' characterize library use as a threat to book sales. Policy implications of this work extend into ongoing public and corporate policy debates about digital content and intellectual property.