Could too much Internet use depress off-line political activity? That’s the intriguing possibility suggested by a recently released Maxwell Poll on Civic Engagement and Inequality conducted by Syracuse University’s Maxwell School of Citizenship and Public Affairs. The national poll of 600 respondents asked citizens about their participation in a variety of politically-related activities over the prior year, as well as the frequency with which they use the Internet as a news source.

The counter-intuitive finding from the poll is that occasional use of the Internet to gather news, as opposed to frequent use or no use at all, is associated with higher levels of political activity. Occasional use is defined as “using the Internet for news only very occasionally.” Frequent use includes “using the Internet for news” either “daily,” “several times a week,” or “several times a month.” Not using the Internet is defined as “never using the Internet for news.” Given these definitions, it seems reasonable to consider citizens within the frequent use category to be different versions of the same civic animal, as opposed to the occasional user or the non-user.

This relation holds up across a variety of political activities, including contacting public officials (by any means), attending political events, contributing money, working in political campaigns, and even voting. Occasional Internet users consistently reported the highest levels of participation; frequent users usually reported the second-highest levels; non-users usually reported the lowest levels.

For example, 46 percent of occasional Internet users contacted a public official in the past year, versus 34 percent of frequent users. Only 31 percent of those never using the Internet made contact. Thirty-one percent of occasional Internet users attended a political event, versus 22 percent of frequent users. Only 12 percent of those never using the Internet attended an event.

It should also be noted that within the frequent Internet use category, participation in off-line political activities goes up as frequency of use increases. For example, 24 percent of those using the Internet several times a month contacted a public official, as opposed to 33 percent of those using it several times a week, and 36 percent of those using it daily. But it remains the case that the occasional Internet user reported higher levels of activity than even the daily user.

One might think that the strange associations between political activity and occasional versus frequent Internet use must be driven by age or gender. After all, we know that political participation increases with age, and that men participate more frequently than women. And indeed, the average age of the citizen who occasionally uses the Internet is 48—considerably above the average age of the frequent Internet user, at 39. But the non-user’s average age is 53, which is prime time for political activity. Recall that non-users were in the lowest category of political participation.

A statistical analysis of all these factors considered together indicates that the association between participation and categories of Internet use holds up, controlling for age and gender. In fact, the association becomes even stronger when controlling for age. Controlling for age, for example, an occasional Internet user is 3.3 times more likely to report having attended a political event relative to someone who never uses the Internet, while a frequent Internet user is 2.1 times more likely.

Controlling for either income or education, however, mutes some of the difference between frequent and occasional Internet use. Under those conditions, both levels of use have strong positive effects on political participation, relative to never using the Internet. Nonetheless, a difference remains. Controlling for income and education along with age and gender, for example, an occasional Internet user is 2.6 times more likely to report having attended a political event relative to someone who never uses the Internet, while a frequent Internet user is 1.8 times more likely.

What’s driving the relationship, if not age or gender? It may be simply an anomaly in the data. No obvious and crisp explanation comes to mind. One possibility is that the occasional use category captures a group of particularly active people who remain more traditional in their orientation toward the Internet as an information source at the same time that it screens out the completely non-active people, who are more likely to be entirely off-line, without a computer in the house.

Another possibility is that while the occasional user occupies the highest category of political activity, overall political activity—across all levels of frequency of Internet use—has been facilitated by the Internet. But it’s also possible that when looking at a cross-section of the entire population, the Internet hasn’t yet significantly reconfigured the playing field. In any event, it’s a strange pattern that invites further reflection and investigation.

In 2004 the Institute for Politics, Democracy, and the Internet (IPDI) published a report based on data collected in late 2003, titled “Political Influentials Online in the 2004 Presidential Election.” “Online political citizens,” defined by high levels of Internet-based political activity, were found to be heavily populated with “Influentials,” a market research term for those who exert a disproportionate amount of influence on their fellow citizens in a variety of areas. While 10 percent of the population falls into this category, 69 percent of “online political citizens” qualified as Influentials.

A potentially important fact to bear in mind when considering the Maxwell Poll findings against those of the IPDI study is that the Maxwell Poll was conducted approximately a year after the 2004 election, and therefore questions about political activity during the previous year refer to a post-election time period. In addition, IDPI focuses its attention on a small sliver of the population, which probably constitutes a subset of my frequent use category.

Finally, a more general point: The Maxwell Poll focuses on using the Internet to consume news, which is a common tack in the growing field of politically-related Internet research—see for example some of the other work by IPDI, and the work by Pew. The idea is that in order to glean the effects of the Internet on political activity, you explore whether exposure to and consumption of Internet-based information is associated with involvement in other political activities.

Of course, close observers of the Internet’s effects on politics know that the dynamic is far more complicated and layered than that. Furthermore, Internet-based news consumption includes many disparate topics, not all of them related to politics or public affairs. So in this case frequent use of the Internet for news captures daily readers of Mariah Carey’s latest doings as well as Vice President Dick Cheney’s. The problem for the social scientist, however, is that asking questions about Internet-based political activities and then comparing that behavior to off-line forms of political activity can quickly become a circular inquiry.

What is really needed to get at the question is an extended qualitative analysis of individuals and groups drawn into political activity through the Internet, which I and some colleagues here at Syracuse have attempted to provide in our own research into the Internet’s political effects. In the meantime, these results from the Maxwell Poll provoke some thought.

Grant Reeher is Associate Professor of Political Science at Syracuse University’s Maxwell School of Citizenship and Public Affairs. He is the author of First Person Political: Legislative Life and the Meaning of Public Service (New York University Press) and co-author of Click on Democracy: The Internet’s Power to Change Political Apathy into Civic Action (Westview).