Unnamed sources

FOR sure, every journalist has used them. In every country that has a free press or even a poor excuse for one, unnamed sources are part of the journalist’s toolkit. In this country, especially, anonymous sources are all over the newspapers and even television, especially the shows that survive on showbiz scandal. In radio, these unnamed sources of juicy information (or better yet, even juicier rumor or gossip), are literally coming out of our ears.

There are, of course, legitimate reasons for using quotes from unnamed individuals or information whose source cannot be attributed. The Watergate investigation, for one, relied heavily on "Deep Throat," whose identity up to now remains the subject of public speculation. It can be argued that without unnamed sources,  significant information that has an impact on politics and policy will never be made public.

But the practice has also been much abused. Unnamed sources are used as shortcuts. Instead of digging up documents, talking to more people, and piecing together information from many sources, reporters often use a  single, unnamed source to make grave accusations of corruption or abuse of power. The race for a juicy headline — on both the reporters’ and editors’ parts — makes the practice a common one. Anonymous sources save newspapers from the risk of being scooped or of letting a day pass without a scandal on page one.

The risk, however, of such wanton use of unnamed sources is that they damage the credibility of journalists. Daniel Orkent, the public editor of The New York Times, says that the most number of complaints he receives from readers has to do with unnamed sources. Readers, he wrote in a recent online column, question the authenticity of unattributed information. They also think that the reporters are being lazy when they quote unidentified individuals. Worse, some readers believe that the unidentified quotes are actually made up by the reporters themselves.

In March 2004, The Times, reeling from a scandal involving an inventive reporter, issued more stringent reporting guidelines. One of these involved unnamed sources: "When we use such sources, we accept an obligation not only to convince a reader of their reliability but also to convey what we can learn of their motivation."

This particular guideline, it seems, has been observed more in the breach. As the reporting on U.S. intelligence leaks on what now look like nonexistent weapons of mass destruction in Iraq clearly shows, the reliance on unnamed sources, especially by those who cover sensitive intelligence, defense, and foreign-policy matters, had damaging effects, both in denting the paper’s credibility (The Times later issued an apology for its reporting on Iraq) and in getting public and official support for an invasion of Iraq.

Okrent says that The Times will soon issue more stringent guidelines in a document called, "Preserving our Readers’ Trust."  In the meantime, he notes that  while the 2004 guidelines reduced by 24 percent the number of anonymous sources used in news stories in the main section of the paper, the percentage of stories citing unnamed sources barely slipped, from 51 percent to 47 percent.  Rarely did the paper provide meaningful description of the sources and the reason for granting anonymity.

After all, newsroom practice, he says, favors anonymous sources. Reporters who get scoops, even if these are based on unattributed information, are rewarded by editors who give more prominent play to their stories.

This problem has been much discussed within the PCIJ as well. One of our most controversial stories, one on presidential paramour Rosemarie Arenas published in 1995, was based largely on unnamed sources. We realized this shortcoming and so made sure that the report: 1) used only first-hand sources (those that knew, saw, or heard information themselves and did not get it from another source); and 2) used only information that was corroborated by at least two other sources.

Our reporting in 2000 and 2001 on the accumulations of then President Joseph Estrada was less controversial in terms of methodology because we were able to dig up a mass of documents in the course of the investigation. In contrast, many of the facts in the Arenas story could not be documented on paper.  That report  had to rely on people — in fact over 50 people, mostly Ramos officials, were interviewed, and we felt that the quantity and quality of the interviews justified the publication of the investigation.

Our 1995 report which examined jueteng in a Pangasinan town was based not just on unnamed sources, but it also did not name the town on which the report was written. This was because the writer used anthropological, rather than journalistic, research methods. It did not matter which town it was, what was being documented was jueteng as a way of life and its rootedness in local politics and culture. To our minds, the use of descriptive research methods to document local knowledge and local "meaning" justified the lack of attribution. After all, the writer was not making a journalistic exposé on a jueteng lord. Instead, through interviews with scores of ordinary residents, local officials, locally influential individuals, even priests, he was describing the persistence of jueteng in a community whose identiy was not central to the story.  His methods were sound ones based on social science research and we had no reason to doubt his conclusions.

There are therefore no clearcut rules for anonymous sources. What is clear to us is that no report should ever be based on a single, unnamed source. Our practice is that unattributed information should be corroborated either by documents or another, preferably named, source. Multiple sourcing is the rule.

Where we have sometimes lacked is in having clearer, or using the Times’ word, more "meaningful" descriptions of the sources that are unnamed. We have also not always been clear about the reasons why the sources were not identified and their motivations for revealing information not explained. Our most recent report on corruption in local health projects, for example, starts off with an unnamed municipal doctor in a government clinic in an unidentified town in Bulacan.

While there was in this case "meaningful" attribution (i.e. a doctor in a rural health unit in Bulacan not just "a knowledgeable source"), should we also have explained why the doctor was not named as well as what motivated her to talk to us? The more discerning reader would assume that the doctor risked losing her position if she were identified. After all, the same doctor is later quoted as saying the municipality had bought overpriced medicines and she was made to sign the vouchers (making her, by the way, a first-hand source).

This report, which reveals that kickbacks from drug procurements by local governments range from 10 to 70 percent, was based on interviews with a number of doctors, auditors, officials, and drug suppliers, nearly all of them unnamed. It was similar in methodology to a report the PCIJ published in 1999 on overpricing in schoolbook contracts in the Department of Education. That report was largely based in interviews with more than two dozen textbook suppliers, all also unnamed.

In both these reports, the sources corroborated each other. Moreover, they presented documents to support their information. In both these reports, the government suppliers were willing to talk because they were tired of the corruption and felt they could conduct business, with much less damage to the public, in a more honest fashion.

In cases like these, should we  leave the reasons for not naming sources to the readers’ discernment? And would it really matter to the credibility of the story if we took trouble to explain why these people were spilling the beans to us? 

The problem is that if we have to explain all these, the narrative gets complicated and the writing becomes clunky. And if several unnamed sources are used in a story, the line "the supplier/ nurse/official, etc. requested anonymity for fear of losing his/her job/contract but agreed to talk because he/she felt the situation had become so bad" becomes needlessly repetitive. Moreover, sometimes it is not clear to us why sources talk. Some clearly have an axe to grind (in which case, we can refer to this in the report). But  many other sources make startling revelations simply because someone asked them in a nice way or because they had long wanted to make these things known. Others talk not because they are unusually civic-minded but simply because they were assured  their identities would be protected. Sometimes, they agree to be interviewed only because  they were friends of the reporter or a second cousin of a former classmate now in the U.S. Once we delve into the motives of our sources, we get into a slippery slope. Motivations should perhaps be disclosed only when they matter to the story and to how the facts revealed would be interpreted.

At any rate, this is too long already.  All I really wanted to say is that sourcing and attribution are never a simple matter and we encourage you to take part in this discussion with this.