Smash hit podcast Serial built audience appeal on questionable reporting practices

So much was impressive about Serial that I’m reluctant to point out some of the intractable ethical problems on which the sensational podcast, which explored over 12 weeks the 1999 murder of a Baltimore teenager, was built. But since Serial’s spectacular success is bound to inspire imitators, and the people who do it next may not be as careful as Serial’s creator, Sarah Koenig, it’s vital to identify these pitfalls, which aren’t minor.

The series, which concluded in mid-December, is widely hailed as the most popular podcast in the admittedly brief history of the medium—some 850,000 downloads per episode—and has spawned virtual boroughs of fans and commentators.

But apart from a posting on the ThinkProgress.org website in November, I haven’t seen any serious meditations on its ethical shortcomings, most of them baked into the core conceit of the program: It aired as a journalistic work-in-progress, sharing its incomplete reporting as it went along.

Serial re-examined the case of 18-year-old high school senior Hae Min Lee, murdered one afternoon in January 1999. Her body was discovered in a park a month later. Police were led to her ex-boyfriend, Adnan Syed, by a friend of his who told police he helped bury the girl. Syed, then 17, allegedly strangled her in a jealous rage and is serving a life sentence.

Koenig, an ex-Baltimore Sun reporter with strong journalistic chops, began looking into the case when she was preparing an item for This American Life, the quirky and compelling public radio series created by Ira Glass. With Glass’s support Koenig and her producers spent months on a minute investigation, interviewing dozens of witnesses, some never contacted for the trial.

That description doesn’t begin to convey how engaging Serial was for its listeners, myself included, and how masterfully Koenig both reconstructed the case and probed the limits of memory and evidence on which her reconstruction was built.

I have nothing but admiration for her craft, but as to whether Serial is a model of “incredibly hard, original journalism” (Philadelphia Inquirer) and a “daring, living piece of serious, journalistic work,” (NBC News), I have big doubts. Most derive from the fact that Serial began airing well before Koenig finished her investigation, and she shared her hunches, speculations, and suppositions as the series unspooled.

That may seem like a commendable reflection of reportorial humility and transparency, but when I first heard Koenig was going public with findings before she had reached any conclusions about what they meant, I was aghast.

First, doing that creates a huge potential for gratuitous defamation if it means sharing damaging conjectures based on suspicions that turned out to be ill-founded. The more the producers advanced hypotheticals about the truthfulness, hidden motives, competency, or credibility of a source—the what-if’s that Koenig presented routinely—the more they put people at risk of unwarranted disparagement.

Any good reporter, conducting an inquiry as empirically challenging as Serial’s, invariably raises just those questions, but keeps quiet about them unless or until they’re determined to be well-founded. But there’s no effective way, two weeks after wondering publicly if somebody is a backstabbing liar, to erase the taint by telling listeners it turns out she was telling the truth.

Second, Koenig was repeatedly going back to a pool of sources her own podcasts had contaminated. She was hearing from informants who had already heard from her and who had overheard what other sources were telling her. This produces something akin to confirmation bias—researchers’ tendency to see only what they’re looking for—but raised to an orange-alert potential for evidentiary mutilation.

It would be impossible to filter out the influence of earlier episodes on informants who either surfaced or resurfaced later on. Nor would it be possible to identify sources who refused to talk because, based on things they heard Koenig say, they mistrusted her or didn’t feel they’d be treated sympathetically.

A third problem also derives from the open-notebook nature of Koenig’s inquiry. It has to do with the identifiability of sources who wanted to stay unnamed. The most prominent was “Jay,” the onetime friend of Syed’s who said he helped bury Lee and became his principal accuser, and went public only after Serial ended in mid-December. But his privacy claim is weak: He was the key prosecution witness; keeping him anonymous was neither feasible nor desirable.

Other fans, however, convened by a spirited Reddit moderator, have had a go at exposing figures peripherally involved in the case, and, in my view, casting a shadow on the confidentially they expected when they came forward. Serial turned out to be an invitation to an outing and, potentially, a warning to future informants. Whether it deterred potential sources from talking isn’t clear, but that’s the danger in a setting where reporting continues amid piecemeal publication.

This is no way to run an investigation. Koenig herself recognized those problems when she told a reporter for The Guardian that she felt as if she was writing a book: “The only difference is if I was writing a book, I would do all the reporting first and then organize it like a normal person and then publish it, but here, I am releasing my work as I go, which is sort of crazy, I realize.”

Serial is over now, and has contributed mightily to the development of podcasting. It was riveting, but it’s a success that privileged narrative brilliance and audience engagement over evidentiary integrity, and relied on story-telling liberties that are incompatible with responsible journalism.

-30-