JOHN KENNETH GALBRAITH once said of Washington journalism: "Nearly all of our political comment originates in Washington. Washington politicians, after talking things over with each other, relay misinformation to Washington journalists who, after further intramural discussion, print it where it is thoughtfully read by the same politicians. It is the only completely successful closed system for the recycling of garbage that has yet been devised."

That was Galbraith's talent for shrewd wit. Now the computers, unfunny and unshrewd, have come up with their version. Stephen Hess, a thoughtful observer of politics, has done the proper modern exercise; he has submitted questionnaires to Washington reporters. using social science techniques of surveying, coding, tabulation and computerized correlations. He comes out disturbingly close to Galbraithhs joke:

Washington reporters tend to get their news from unidentified political sources who leak material for their own benefit.

They tend to produce news centered on surface events and personalities, and think of documentary research and substance as "boring."

Washington reporters are increasingly conservative or apolitical, more elite and more detached from the realities of American life.

The reporters know the worst about themselves. They (82 percent) agree that pack journalism is a problem; that there is insufficient coverage of important but less glamorous beats like regulatory agencies (58); that Washington reporters make to too easy for sources to hide their identity (67 percent); that they concentrate excessively on breaking news (51 percent); and that they are out of touch with the country (82 percent).

They go along with this flawed pattern. Hess's responses show, for reasons that include aoivdance of risks by running with the herd, the attraction of the pay and prestige, and because, like the politicians they cover, they like the instant gratification of quick stories.

The product of this flawed system seriously influences all American news, Hess reports. The reporters' home organizations use growing quantities of this national news at the expense of important local and regional news. They do it partly because the editors perceive their audiences as pre-conditioned to want this by television, and partly because using as much as possible of the news from their fixed-cost Washington operations and services is cheaper than live local and regional coverage.

Hess identified 1,250 Washington reporters (he discards the traditional Potomac honorific, "correspondent") who cover the national government for commercial news outlets. He submitted a 16-page questionnaire to each one, conducted 150 personal interviews, 194 random telephone interviews, talked to news executives in six cities, and got a net of 476 usable responses, representing 38 percent of Washington reporters. In addition, he analyzed a week of Washington stories in 22 newspapers, three network news shows and a newsmagazine, coding 2,022 stories in order to see what happened to Washington stories when they reached home outlets.

Some results are not surprising. The Washington press corps tends to be "male, white, urban and Northeastern" and the top bureau people tend to come from elite private universities and dominate coverage of prestige beats like diplomacy ("one elite reports on another elite").

But some Washington journalists, editors and politicians will be surprised.

The overwhelming preference for reporting quick surface events betrays the picture of a horde of liberals obsessed with investigating the establishment. Curiously, most reporters said they thought most other reporters had a liberal bias. But, Hess says, "perception has not caught up with reality." There is a new main body of reporters in Washington. Most of them -- 58 percent -- are in their twenties and thiries, and they are largely uninterested in political ideas and mainly concerned with excitement, prestige and career.

"An apolitical press corps in Washington, if this is the case, looks like a strange finding," Hess says. "One might expect that those covering the national government would have a deep interest in political ideas. Instead the lure of Washington seems to be otherwise; the attraction perhaps is to excitement and powerful personalities."

The continuing concentration on surface events and color at the expense of substance represents a lack of intellectual depth and sensitivity, Hess suggests. He thinks that at least the journalistic leaders like editors and bureau chiefs ought to have some intellectual distinction. But that is not how journalistic leaders are selected, he says.

This is a careful, properly statistical and cautiously analyzed update of Leo Rosten's similar study in 1937. But having completed the obligatory tour into "precision journalism" that tries to quantify the human scene, I hope that Hess, with the meticulous printouts now integrated into his subconscious, will use his admirable intuitive powers to produce an extended work on what this means for the perceptions and misperceptions that Americans have of their political system.