The president of the United States, both an avid consumer and a vicious antagonist of news, will in one breath vilify the (FAKE NEWS) media and in the next praise Fox for its ratings.

Americans’ trust in media fell last fall to its lowest point since Gallup began polling on the issue in 1972, driven in large part by growing distrust from Republicans. But while a slim percentage of Americans regard “the news media” in abstract as trustworthy, when asked specifically about news outlets they consumed most often, more people had favorable views, according to a new study released Wednesday from the Media Insight Project (a collaboration between the American Press Institute and The AP-NORC Center for Public Affairs Research).

Just 17 percent of Americans said they found “the news media” in general to be “very accurate,” but asked specifically about the outlets they relied on for news, that number doubled to 34 percent.

Similarly, just 24 percent of Americans said they regarded “the news media” as “moral,” but that number jumps to 53 percent for the media they consumed most often.

58 percent of those surveyed also felt that the news media as a whole tries to hide mistakes. They were more forgiving of media they used most often — 32 percent felt their news media of choice try to cover up mistakes.

If this phenomenon seems familiar to any political science majors out there, it resembles Fenno’s Paradox, an idea political scientist Richard Fenno outlined in a 1972 lecture about Congress. Broadly speaking, people hate Congress — but they tend to like their local member of Congress. While approval of both the institution and the individuals can wax and wane, the preference for your local representative is consistent. Just as people who hate “Congress” also reliably reelect the vast majority of its members every two years, a disdain for “the media” can coexist with an approval — even affection — for the news outlets one chooses to take in.

As Fenno wrote (try mentally subbing in the media equivalents to Fenno’s legislative subjects):

…we apply different standards of judgment, those that we apply to the individual being less demanding than those we apply to the institution. For the individual, our standard is one of representativeness — of personal style and policy views. Stylistically, we ask that our legislator display a sense of identity with us so that we, in turn, can identify with him or her — via personal visits to the district, concern for local projects and individual “cases,” and media contact of all sorts, for example. On the policy side, we ask only that his general policy stance does not get too frequently out of line with ours. And, if he should become a national leader in some policy area of interest to us, so much the better. These standards are admittedly vague. But because they are locally defined and locally applied, they are consistent and manageable enough so that legislators can devise rules of thumb to meet them…

For the institution, however, our standards emphasize efforts to solve national problems — a far less tractable task than the one we (and he) set for the individual. Given the inevitable existence of unsolved problems, we are destined to be unhappy with congressional performance. The individual legislator knows when he has met our standards of representativeness; he is reelected. But no such definitive measure of legislative success exists. And, precisely because Congress is the most familiar and most human of our national institutions, lacking the distant majesty of the Presidency and the Court, it is the easy and natural target of our criticism. We have met our problem solvers, and they are us.

Back to today’s study: Consistent with other similar research, a strong partisan divide persists when it comes to Americans’ feelings on the role of the press in holding politicians accountable, and whether the press treats “all sides fairly.” The gap on perception of fairness was less pronounced when the researchers asked specifically about media they consumed.

Americans are also divided on what exactly they’re counting as part of “the news media.” Just 5 percent of Republicans said they considered websites like Politico or Breitbart as part of “the news media,” yet when asked specifically about news outlets they used, 16 percent identified those sites as regular sources of news for them. These sorts of mismatches, the study suggests, “show how each party can have a conception of ‘the news media’ separate from the news media they use.”

The findings were based on a survey of 2,036 Americans — most of them active news consumers — half of whom were asked about “the news media” in general and half of whom were asked the same set of questions about “the news media you use most often.”

The full report, which includes additional analysis broken down by age and ideological leaning, is available here.