one of the other two sites, and only 32% of these stories popular in more than one site dealt with politics, business, economic, and international matters. This amounted to a 16 percentage point gap between the levels of similarity of the top news choices of journalists and consumers, and a 27 percentage point gap between the thematic composition of these similar choices.

ILLustratIon By raymonD BIesInger

That initial study begged a follow-
up question: do similar patterns arise
elsewhere? A second study showed
that the mismatch also applies to the
leading, elite media in the U.S. My col-
laborator Limor Peer of Yale University
and I conducted a study that compared
the concurrent news choices of jour-
nalists and consumers of four leading,
U.S.-based sites: CNN, Yahoo News,
Chicago Tribune, and the now-defunct
Seattle Post-Intelligencer. 3 We chose
these sites to represent broadcast, on-
line-only, and newspaper parent com-
panies, and also different geographic
orientations. The first two sites were
national-global, while the second two
were local.

We worried that supply and demand are interdependent: journalists might prioritize certain stories because they perceive that consumers could find them appealing, and consumers might click more often on stories that receive major treatment by journalists. To try to get past these interdependencies, we conducted a second analysis on each site that excluded the stories that were selected by both journalists and consumers. Focusing on the stories that journalists chose irrespective of their level of popularity among consumers, and those that consumers chose even though they were not prominently displayed on the homepage, would give us a stronger measure of each group’s independent preference.

The results suggest that independent from the influence of each other, the choices of journalists and consumers would follow strikingly diver-