Even though Liberals and Republicans have been complaining about the unfairness of the press reporting on their candidate, recent research by the Pew Research Center’s Project for Excellence in Journalism shows a more complex picture where the level of negativity tracked the voters’ perception of the candidates.

The conclusions are well worth repeating

What we see in these findings, above all, are two phenomena. The first is the focus on tactics and strategy. The candidate who was perceived to be winning this year got better coverage. We have seen that pattern before. In 2000, our research saw George Bush receiving more positive coverage than Gore. In 2004, our studies of a narrower time frame saw Kerry enjoying better coverage, as polls perceived his closing the gap on Bush.

The second phenomenon is an almost instantaneous reinforcing and echoing effect of the press. Presidential elections are now so heavily polled, with various daily tracks and compilations of state-by-state polls, that every campaign event is almost instantly measured for its political impact and that in turn is immediately analyzed by the political press. Each event has in a sense three echoes. The event is covered. The effect is measured. And the reaction to that measurement by the campaigns is then examined and covered.

That pattern becomes a snowball, and the trajectory of any one campaign event is magnified.

Occasionally the news media will alter this dynamic pattern of covering the candidates’ behavior and then measuring the political effect of it—with their own enterprise or background pieces. This year, stories about Palin’s record or Obama’s relationship with a former terrorist are examples. These may in turn influence the behavior of the candidates. But these enterprise stories represent a fairly limited portion of the coverage. And that enterprise may be diminishing as the press suffers cutbacks in resources. To that extent, the agenda-setting influence of the media may be diminishing.

Even more occasionally the press will influence the narrative by asking questions of the candidates directly. This, however, is rare in presidential general election periods, when candidates severely limit their exposure to the traveling press corps for direct inquiries. There is relatively little of that in this sample, save for the Palin interviews.

The story of the media coverage of the general election phase of this race, at least so far, is the snowball effect on John McCain. In the press, this race became substantially defined around the troubles that began for McCain in mid-September and the difficulties he had in trying to change the trajectory of the race in the weeks that followed. With a press corps heavily focused on horse race, and a candidate whose strategy was apparently thrown off by events, he has been unable to change that narrative.