Among other things, the study looked at stories to determine who predominantly they impacted by the way stories were put together. Who was affected by the events described, and when the implications of events were analyzed, was it the implications for citizens, for the politicians, the parties, interest groups or others?

During the two weeks studied, the vast majority of stories, 73%, were produced in a way largely oriented to how politicians would be affected. Only 20% were written in a way that predominantly explained the impact on citizens. This number is down from 27% four years ago.

These findings may be discouraging to those who have worked for that last decade to make political coverage more voter oriented. The effort to redirect political coverage more toward the concerns of citizens apparently has not significantly influenced the way coverage is constructed. The trend, rather, seems to be moving in the opposite direction.

The focus on politicians was even greater in debate stories. More than 90% of debate stories were written largely about the impact on the Kerry and Bush campaigns, up from 74% of politician-centered stories four years ago. Only 8% made clear how citizens might be affected by what was being described, by outlining, for example, the implications on citizens of the candidates' policy proposals. That number was down from 15% in 2000.

Some may argue that focusing on impact of debates on the politicians is appropriate given that historically these events imply high risk for candidates who perform poorly. Others, however, may point out that debates are also moments when voters are known to pay more attention, and get a clearer sense of a candidate's positions and proposals. This year the debates generated higher audiences than four years ago.