This study investigates how information and communication technology affects civic engagement in politics and policy processes. It develops a typological framework of information and communication technology (ICT) use, which posits that individual ICT applications have different implications for civic engagement, and that individual engagements should be differentiated from one another. Then it employs the recursive bivariate probit model (RBPM) and the propensity score matching (PSM) method to analyze the 2004 Pew Internet Tracking Survey data.

Findings suggest that e-government use, a type of ICT use, has a significant positive impact on nonpartisan civic engagement (sending emails about the campaign and voting), but moderate or negligible effects on partisan engagement (i.e., solicitations for votes, financial contributions, and attendance at a campaign rally). In contrast, use of campaign websites significantly influences both partisan and nonpartisan engagement.

The RBPMs of nonpartisan engagement have significant endogeneity, which indicates the need for consideration of the indirect effects of such information technology factors as Internet experience and online use intensity. The effect of e-government use is direct and exogenous in partisan engagement. PSM produces a robust average effect of e-government use, which is smaller in partisan engagement than in nonpartisan engagement. Use of campaign websites is influential, especially for nonpartisan engagement and financial contributions.

E-government, largely used in the “economy model,” is less influential than campaign websites but appears to have great potential for getting ordinary citizens engaged through its growing emphasis on public outreach. Campaign websites in the “solidarity model” are primarily designed to mobilize party members and supporters rather than the general public. E-government appears to support the optimistic “transformation theory,” while campaign websites uphold the pessimistic “reinforcement theory.”

The typological framework helps people understand the way that ICT use influences civic engagement in the public sphere. This study explores key issues such as causality, endogeneity, simultaneity, selection, and the missing data problem and thus improves the methodological rigor of research in this field. However, this study is limited by the survey data set since it does not contain sufficient variables for various civic engagements and ICT applications.