Posted: April 18, 2000
This page is categorized as:    link to Crisis Communication index
Hover here for
Article Summary This was the first “crisis” I watched unfold. The themes have since become familiar (they are playing out now in the terrorism crisis): The sources minimize the risk and over-reassure the audience; the experts are uncertain and disagree with each other; the public doesn’t panic but everyone keeps thinking it will; the color stories overpower the technical stories. (Writing this article in 1979 changed my career; see “Muddling My Way into Risk Communication.”) Includes two sidebar articles: “The local media feel the heat” and “The Inquirer goes for broke.” For a more recent perspective, see “Three Mile Island – 25 Years Later.”

At Three Mile Island

Columbia Journalism Review, July/August 1979, pp. 43–58

На Македонски:   На островот Три Миꙥји 
(posted in March 2012 on

It was a tough assignment. Some reporters called it the toughest they ever had. Two journalists who went to cover the press found fear, confusion, and some very good work.

We walked into the Harrisburg capitol newsroom Saturday evening sporting radiation badges, the lithium fluoride kind – T.L.D.’s to those now versed in radiation terminology. When we got home a radiologist would process the badges and tell us our dose. The first person we met was a Los Angeles Times reporter, one of eight sent to cover the accident at the nearby Three Mile Island nuclear plant. “Are you wearing those sissy things?” he said with a smirk.

Nuclear macho. We saw plenty of that in the next few days. But in the next few hours we saw fear, chaos, and anger instead. The Associated Press was about to report that a hydrogen bubble inside the T.M.I. reactor core could explode.

Until now, the excitement at Three Mile Island had come in surges. The day it all began, Wednesday, March 28, was dominated by the slowly leaking radiation, and the slowly leaking news that despite the denials of the utility, Metropolitan Edison, the accident was probably the most serious in the history of the American nuclear industry. RADIATION SPREADS 10 MILES FROM A-PLANT MISHAP SITE was the headline over Tom O’Toole’s story in the next morning’s Washington Post. Thursday the tension subsided; Met Ed, the Nuclear Regulatory Commission, and Pennsylvania Governor Dick Thornburgh agreed that the crisis was past, and editors who had staffed the accident quickly and heavily began wondering if they had overreacted. Bryce Nelson’s story in the Los Angeles Times carried a one-column head: PLANT STILL LEAKS RADIATION BUT PERIL SEEMS OVER. Friday all hell broke loose, with two more uncontrolled bursts of radiation, a recommended evacuation of preschool children and pregnant women, and a mysterious bubble that some N.R.C. officials said might be hard to remove without risking a catastrophic meltdown. Friday night, Walter Cronkite intoned that “the world has never known a day quite like today,” and The New York Times put four Three Mile Island stories on Saturday morning’s front page, under this foreboding headline: U.S. AIDES SEE A RISK OF MELTDOWN AT PENNSYLVANIA NUCLEAR PLANT; MORE RADIOACTIVE GAS IS RELEASED. By Saturday more than 300 out-of-town reporters had converged on Harrisburg.

But Saturday was quiet there. Met Ed and the N.R.C. continued to disagree over the size of the bubble and the seriousness of the accident, but the big news came out of Washington. N.R.C. Chairman Joseph Hendrie suggested that if simpler methods didn’t work, a twenty-mile-radius mass evacuation might be a good idea before they really went after the bubble. The N.R.C.’s man on the scene, Harold Denton, hadn’t talked to reporters since the night before, when he had wryly admitted that the bubble was a “new twist” not envisioned in the commission-approved emergency procedures. Crowded into a spare room in gubernatorial press secretary Paul Critchlow’s capitol office, the N.R.C.’s public-information people had little to add besides a list of the maneuvers Denton might try to get rid of the bubble.

At 8 p.m. Saturday there were only twenty reporters in the capitol newsroom, most of them complaining to each other that Critchlow was unreachable and that his staff wouldn’t say whether the lid was on for the night or not.

Except for the newcomers like us, they looked haggard. Three Mile Island was a tough assignment; some veterans said it was the toughest they’d ever had. Somehow the TV reporters managed to stay crisp, but the print people showed the strain as they slogged through one sixteen-hour day after another. They posed a problem for capitol police guard Tom Chiricos. With so many journalists crawling around, police had suspended the policy of checking credentials at the door. “They told me that if they don’t look like derelicts to let them in,” Chiricos explained. “But a lot of them do look like derelicts.”

Jim Panyard, a veteran Philadelphia Bulletin correspondent whose seniority and centrally located desk made him unofficial newsroom host, had sent his family out of the area Thursday morning. “I work here,” he said matter-of-factly. “If I left I’d lose my job.” Panyard was rumpled, tired, tense, and angry. He could cover a legislative session in his sleep, but this was unlike anything he had handled before. Sources seemed to speak a foreign language, he said with a sigh. You asked them a straight question about how much radiation is escaping and they answered with mumbo-jumbo about millirems, manrems, rads, and picocuries. Once you had figured out what they were saying you discovered another source was saying something different – and without a nuclear physics degree you couldn’t come up with the right follow-up question to tell who was lying. “We’ve been given complete misinformation and conflicting statements from the N.R.C. here, the N.R.C. in Washington, the governor’s office, and the utility,” Panyard said. “There is no doubt that the situation is dangerous, but how dangerous is the question. I’m concerned, and I think other reporters are, too.”

Two reporters loosened up by hurling a frisbee the length of the newsroom, past a wall of framed portraits of past capitol correspondents – journalists from an era when reporters didn’t have to know about fuel rods and hydrogen-oxygen ratios to ask an intelligent question. A bulletin board on the opposite wall bore a message to out-of-town reporters, advising them that if they noticed the streets of Harrisburg were dark and empty at night “this is not – repeat not – because of the accident at T.M.I. That’s the way they always are.” It was signed “the locals.”

Other reporters passed the lull refining their gallows humor:

  • Hershey, Pennsylvania; it melts in the ground, not in your hand.
  • I went to bed last night and turned out the lights – but the room didn’t get dark.
  • Weather report: It’s partly cloudy out, with a 40 percent chance of survival.

The Troublesome Bubble

It was four hours before Sunday, April l, and the A.P.’s report that the bubble could explode was about to move out of Washington. To many reporters in Harrisburg it would later look like an ill-conceived April Fools’ joke. Most newspaper readers would never see the story, for the next few hours of journalistic frenzy would force the A.P. to modify it well before deadline for Sunday’s late editions. But for Saturday night TV viewers around Harrisburg, and for reporters in the capitol newsroom, the story would heighten fears dramatically. (It would also duplicate in many essentials a more widely discussed U.P.I. story of the day before, which reported that an N.R.C. technical expert in Washington had admitted that trying to reduce the bubble could possibly cause a meltdown. U.P.I.’s bulletin had provoked a near-panic in Harrisburg and had prompted a carefully worded N.R.C. release which, while acknowledging the possibility in forbiddingly technical terms, seemed to deny it.)

The A.P.’s story Saturday night was a genuine scoop, one of the few to come out of Three Mile Island. Apart from this, it bore all the hallmarks of Three Mile Island reporting during those first few hectic days. Journalists were overtired and fearful. Sources and public-information people were hard to reach. When reached, they gave out conflicting stories. And it turned out they were all guessing.

The story begins in the A.P.’s Washington bureau, where special energy writer Stan Benjamin had a tip from an N.R.C. source that the bubble was in trouble. Not only was there danger in removing it; there was danger in leaving it alone because oxygen might well be seeping into the hydrogen, creating a mixture that could flame or explode. Benjamin called N.R.C. spokesman Frank Ingram. Is there oxygen in the bubble? Yes. Could that produce an explosion? Yes. What would happen? Possibly it would cause a meltdown; more possibly it would blow the head off the reactor; certainly it would do damage. How soon? Maybe several days. The timing was crucial, but Ingram wouldn’t be pinned down. Then A.P. reporter Andy Schneider, on loan from the Concord, New Hampshire, bureau where he had covered the reactor- siting controversy at Seabrook, called a source at the N.R.C. “Two days,” the source said.


“As soon as the advisory hit the wire, I got a call from Ingram,” Benjamin recalls. “‘What have you guys got out?’ he asked. ‘I heard you put out something that said the bubble is going to explode.’ I read him the advisory and he said there was nothing wrong with it.” Benjamin had chosen his words carefully. “I didn’t want it to get hyped. Ingram had told me there was ‘a possibility of the bubble becoming explosive.’ I wrote that it ‘shows signs of becoming potentially explosive.’ I qualified it three ways.” The story that hit the radio wire at 8:50 used the same phrase.

A matter of days

Back in Harrisburg, Panyard paced the capitol newsroom, shaking his head over a copy of the radio story. “A.P. says the bubble is unstable,” he announced, clearly upset by the news. In the lounge next to the newsroom, a TV set that nobody had been watching now carried CBS and NBC bulletins about the bubble, somehow sharpening the sense of imminent disaster. Someone called Lombardo’s, where reporters were grabbing a late dinner; they stuffed uneaten food into doggie bags and rushed back to the capitol. About twenty dashed up three narrow flights of stairs to the governor’s press office, demanding confirmation or denial. “They burst in,” Critchlow recalls, “and said ‘What the hell is going on? We want to know if we should get out of here.’” They weren’t after a story; they wanted to know if they were in danger. Critchlow assured them there would be no explosion that night.

By 9:02 Benjamin had filed a rewrite of Tim Pettit’s earlier A.P. story from Harrisburg, this time leading with the new information that he and Schneider had discovered. Once again Ingram called, listened to the story, and approved it. Five minutes later he called a third time and asked Benjamin to read the story to N.R.C. deputy director Edson Case. “Ingram said again that it was okay. I asked Case if the story was correct and he agreed. I confirmed that story three times in addition to the original interview,” Benjamin says.

In Harrisburg. meanwhile, Critchlow waylaid Denton on his way to brief the governor and detoured him into the newsroom, where about forty reporters had gathered. It was now 9:30. Critchlow was irritated. Denton was calm, wearing his perpetual cowboy grin. He was accompanied – as always – by N.R.C. information head Joe Fouchard, chewing on an unlit cigar.

“Please,” Critchlow pleaded, “let’s put this story of the hydrogen explosion to rest.” As nervous reporters pelted him with questions, Denton reeled off a series of figures. The bubble was 2 percent oxygen; it could become flammable at 8 percent, explosive at 16. Extrapolating the rate of accumulation since the accident, he figured they had twelve days before an explosion was possible. But he did not intend to let the bubble “just sit there,” Denton explained with his engineer’s can-do optimism. If he couldn’t shrink it one way, he would try another.

The Harrisburg A.P. bureau called Benjamin in Washington. “Denton denied your story,” they told him, repeating Denton’s remarks. “That doesn’t sound like a denial to me,” Benjamin countered. “He just thinks it’s going to take longer to become explosive.” But by now Denton was the only reliable source on the scene, the White Knight helicoptered in by the White House Friday at Governor Thornburgh’s request, and most reporters believed him. The A.P. filed another story using all three time estimates. U.P.I. fired off an advisory misleadingly telling editors that Denton had said “there was no danger of a hydrogen explosion.”

A little after eleven Denton repeated his statement at a formal news conference with Thornburgh in the governor’s briefing room, this time before nearly 200 reporters and a dozen TV cameras. Thornburgh was peeved at the press; reporters were angry. “It would seem today that the contradictory statements are coming from within your own agency,” one shouted. “I think the contradictions have been overplayed,” Denton replied. “We are in full agreement and constant communication, but somehow when we brief people it comes out different.”

“I hope CJR gives A.P. heat for blowing tonight,” suggested one reporter on the stairs. “But you should also mention,” said another. “that half the reporters here are relying on A.P. for the facts. They did a good job till tonight.”

A rapid change

Benjamin thinks that the A.P. did a good job Saturday night as well. The important news, he says, was that the N.R.C. admitted the possibility of an explosion; the two-day figure was never hard. On Sunday the A.P. dropped it and went with Denton’s estimate, by then revised to five days.

But to reporters (and residents) in Harrisburg Saturday night, there was all the difference in the world between two days and twelve days, or even between two and five. It meant they didn’t have to pack and run. Fear turned to anger, and reporters turned theirs on the A.P. story. In a local watering hole after yet another sixteen-hour day, they muttered about wire services in general. “If U.P.I. has panic, A.P. has got to have hysteria,” declared Ben Livingood of the Allentown Call-Chronicle. Others countered that Denton might be overly optimistic, trying to avoid panic, and that the A.P.’s anonymous source might have been right.

In retrospect it is clear that the N.R.C. simply didn’t know how long it might take for the bubble to explode. The calculations were wild guesses at first, and careful guesses after a while – but still guesses. The experts weren’t absolutely sure there even was a bubble; the hydrogen might have been dispersed through the cooling water in a thoroughly nonexplosive froth. As they unsteadily piled inference on inference – all based on erratic readings taken from instruments not built to withstand the fierce radiation – they paused from time to time to brief the N.R.C. brass. Once in a while the experts or the brass briefed reporters, who complained bitterly that the bubble’s deadline seemed so much less firm than their own.

By Monday morning the bubble, mysteriously, was smaller. Met Ed’s George Trotter announced that it was essentially gone and that “there are no problems left.” (Having learned to distrust the utility, the Harrisburg Evening News put MET ED SAYS in seventy-two-point type over its first edition story.) At his 11:15 briefing, Denton estimated the bubble’s size, 850 cubic feet a few days before, at 50 cubic feet now. “I don’t want to be stampeded into concurring that the bubble is actually this small,” Denton cautioned. “I didn’t expect such a rapid change, and that’s one reason I want a careful look at it.” Was Met Ed distorting the truth again when it said the bubble was essentially gone? The N.R.C.’s Roger Mattson answered for Denton: “The point I’m trying to make is there is not a clear line between here and gone.”

Days later, reporters were still complaining about the A.P. story and misunderstanding its content. Newspaper Enterprise Association columnist Bob Walters wrote that the account was “inaccurate” and “erroneous.” The Philadelphia Inquirer, in a long April 5 wrap-up, mistakenly said that the A.P. had reported that “the bubble is so volatile that it may explode at any minute.” Bill Hoop, who ran the Harrisburg U.P.I. bureau during the crisis, smugly reminded us that they hadn’t moved any inaccurate stories. (Despite Saturday night, reporters were nearly unanimous in praising the A.P. as faster and more thorough than the competition; it had thirty-two reporters on the story, more than twice as many as U.P.I.)

In the next month Benjamin was to discover that his story was entirely accurate – and (perhaps) entirely false. The N.R.C. transcripts revealed that the commission had in fact been extremely concerned about a possible bubble explosion: the A.P. had scooped everyone with that crucial information. On May 1, however, Benjamin listened while Mattson informed the N.R.C.’s Advisory Committee on Reactor Safeguards that a bubble explosion had been impossible from the start. The staff “fouled up,” Mattson told the meeting. No oxygen was being generated within the reactor after all. “The explosion story may have been blown out of proportion by the press,” Mattson said, “but it originated with our staff.”

The One That Got Away

The most important fact about the bubble saga – more important than the conflicting stories or the bewildered and fearful reporters – is that the experts were confused in the face of a crisis they never expected. Indeed, the confusion of the experts may well be the most important fact about Three Mile Island generally. For days the people in authority had little idea of what was happening, and less of what would happen next. They fixed the reactor the way a mechanic fixes a car – they tinkered. When it was all over they tried to figure out what they had done right. With two months’ hindsight, we now know what the big story at T.M.I. really was: no one knew enough to guarantee that the genie would stay in the bottle. Call this criminal incompetence or call it the human condition. Either way it was the story.

Most reporters missed it.

More precisely, most reporters skirted it in what they wrote, although they often suspected it and some of their readers and viewers may have deduced it. Only in mid-April, after the N.R.C. had reluctantly released the transcripts of its frenzied consultations, did stories finally focus on the experts’ confusion. Tom Reid and Ward Sinclair’s April 13 lead story in The Washington Post, for example, said the transcripts “portray an agency that, two days after the Pennsylvania nuclear accident, still lacked any clear idea of how to deal with the problem.” Robert Schakne reported on CBS that the transcripts showed that N.R.C. members “were far more confused about the nuclear disaster at Three Mile Island than they’ve publicly admitted.” David Hoffman, Frank Greve, and Richard Ben Cramer wrote in The Philadelphia Inquirer that the N.R.C. was “still poorly informed, mired in internal confusion” four days after the crisis began.

To understand why this story went largely uncovered for two weeks, think of the T.M.I. coverage in two phases: the Cover-Up phase, focused on conflicting stories from Met Ed and the N.R.C.; and the White Knight phase, focused on Harold Denton’s incredible credibility. The A.P.’s Saturday night bubble was the watershed between the two.

Local reporters got on the T.M.I. story early Wednesday morning when routine police checks revealed that a “general emergency” had been declared at the plant – meaning that radiation was escaping beyond its borders. They rushed to the site, where they found plant workers milling around and police barring the gate to the island. Bob Grotevant of U.P.I. says the Met Ed people were tight-lipped, refusing even to give him company pamphlets on display at the visitors’ center. Standing outside the center, whose telescopes pointed out across the Susquehanna at T.M.I.’s four monumental cooling towers, Met Ed’s Bill Gross – in calmer times a tour guide – announced that there had been a minor accident and the company was coping with it. Gross wouldn’t answer questions. “There was obviously something wrong,” Grotevant recalls, and he was suspicious. But his initial stories, like most, simply reported the statements of the police and the utility. There was nothing else to report until afternoon, when more radiation escaped and the state government started questioning Met Ed’s optimism. Then the conflicting statements began, and reporters began looking for the cover-up.

Reporters assume that sources know. The sources may tell the truth or they may lie, mislead, or stonewall – but they know. Our society, furthermore, assumes that scientists and engineers know. We may not understand what they’re talking about, but they do: ignorance and incompetence are seldom newsworthy, anyhow. These were some of the assumptions reporters brought with them to Three Mile Island.

Dim light from Met Ed

Until Denton arrived Friday afternoon, Met Ed was the main source of technical information. And Met Ed steadfastly understated the accident, alternately denying that there was a crisis and asserting that it was over. Interviewed by his hometown Reading Eagle several weeks later, Met Ed president Walter Creitz acknowledged that the company “should have been more pessimistic,” but denied the cover-up charge. Blaine Fabian, the company’s public relations head, told us “it wasn’t like a train wreck – the facts kept changing from moment to moment. Reporters didn’t seem to understand that.” An A.P. chronology of Wednesday’s conflicting claims left open the possibility that Met Ed simply didn’t know what it had on its hands, but implied that it surely knew more than it told: “Who knew there was a leak? When did they tell? … Utility officials and government authorities didn’t waver from assertions that there was no danger to public health or safety. Little else remained constant as the story unfolded.”

Met Ed had a reputation for defensiveness even before the accident. Last summer Harrisburg magazine ran a fantasy imagining a T.M.I. meltdown; Creitz reacted by writing local officials, urging them to yank the magazine’s C.E.T.A. funding. When the nearby York Daily Record did a series on safety hazards at T.M.I., Creitz again cried foul; two days before the accident the Record duly published his rebuttal accusing the paper of “yelling fire in a crowded theater.”

The company’s credibility died fast. Ten hours after the accident, Lieutenant Governor William Scranton was already telling a news conference that Met Ed “has given you and us conflicting information.” By Saturday N.R.C. and state government people were openly advising reporters to ignore the utility; and the White House called the same day, instructing the company to refer all questions to the N.R.C. (Fabian says the company had already decided to stop commenting.) The strongest evidence that Met Ed had squandered its credibility is that reporters didn’t object to the muzzling – on a story that had all too few sources already.

The relative weight of incompetence and dishonesty in Met Ed’s early statements is hard to determine even now. Journalists, in any case, focused on the cover-up. Thursday morning, in the first of several critical editorials, the Harrisburg Patriot wrote of “an attempt to avoid alarming the public by silence and secrecy, which is the best way to awaken the darkest fears of people.” Roger Witherspoon of The Atlanta Constitution was less delicate. In a weekend press piece he wrote that Met Ed “basically said nothing, contributed little, misinformed N.R.C. officials, misled public officials, and finally shut down because no one believed them anyway.”

Met Ed was responsible for most of the conflicting stories during the first three days, but the N.R.C. in Washington and the state government in Harrisburg contributed their share, especially on Friday. That night the Associated Press moved a chronology of the day’s disagreements that was praised by nearly every reporter we talked to. It began this way:

Minute by minute, the official explanations of Friday’s trouble at the Three Mile Island nuclear plant kept changing: The new release of radiation was unexpected. No it wasn’t. Yes it was. We’re considering evacuating. There’s no need to leave. Some people should go.

From utility spokesmen, federal government representatives and state authorities: Conflicting reports.

The New York Times ran a similar piece by Ben Franklin, complaining about “three days of conflicting and sometimes flatly contradictory statements about the nuclear emergency.”

Reporters usually flourish when sources disagree, playing one source against another and all sources against their journalistic instinct. But T.M.I. was too dangerous for reporters to enjoy the confusion, and journalistic instinct couldn’t help them decide who was right about how much radiation was released or when the hydrogen bubble might explode. In a Friday editorial The New York Times pleaded with the authorities at Three Mile Island to coordinate their stories. “I don’t agree with the editorial,” Charles Mohr of the Times told us the following Tuesday. “Nothing should be done to cut off any source.” But Mohr was an exception. Generally speaking, most of the reporters on the technical story desperately wanted the experts to speak with one voice. When the experts didn’t, they were angry and went after the discrepancies, in the words of Jim Holton of NBC radio, “like petulant children playing D.A.”

Enter Denton

On Saturday the N.R.C. did what the Times had asked. Harold Denton became the point man for all information about the accident; Met Ed clammed up and the N.R.C. in Washington tried to do likewise. At Denton’s first formal news conference Friday night a reporter popped the crucial question: “Has a situation similar to this ever been dealt with before? Do you know what you are doing?” Denton answered hesitantly. “Well, I think we know what we are doing, yes, but we have never had such extensive fuel damage in the life of any reactor.” Later he added that the bubble and the escape of radiation from the containment building were also unexpected. These were the three most dangerous developments to occur at Three Mile Island.

But Denton’s calm demeanor and constant optimism were as reassuring as his frankness was refreshing, and the White Knight won instant credibility. In their eagerness to invest him with authority, reporters called him Doctor Denton, although he has only a bachelor of science degree from North Carolina State. They wrote thumbnail sketches that stressed his jowly grin and unflappable style. But they didn’t write a lead like this: “The N.R.C. and its nuclear consultants from around the world are still unable to state with confidence what is happening inside the crippled Three Mile Island reactor.” The A.P. came closer than most, five days after the accident, with a story that began: “In an industry devoted to stringent safety features and voluminous contingency plans, the dangerous bubble lodged under the roof of the Three Mile Island nuclear reactor is one emergency the emergency planners did not foresee.” The Los Angeles Times ran the story on page nine; few major papers used it at all.

Saturday night’s A.P. exclusive was the last important discrepant account – and reporters at the site believed Denton. “He’s disarmingly frank,” murmured Tony Mauro of Gannett after Denton visited the capitol newsroom to deflate the story. “He can probably be believed that the bubble is stable.” By then the possibility of meltdown or explosion seemed far more newsworthy than either the N.R.C.’s continuing confusion or the diminishing conflict among sources. From Sunday on, the accident as seen through the eyes of Harold Denton dominated the coverage. Only the color stories continued to quote local residents who felt they weren’t getting the truth. Reporters, at last, felt they were.

Only once after Denton came on the scene was the N.R.C. itself accused of lying to the press, but Denton himself emerged unscathed. Late Monday night, parked directly across the Susquehanna from the plant, Rod Nordland of The Philadelphia Inquirer’s investigative team fooled with his fancy scanner radio, searching for T.M.I. transmissions. Nothing on the utility band, nor the police band. He switched to a frequency the instruction booklet said was reserved for “federal interagency cooperation during a nuclear war.” And there they were.

“There’s a direct leak from the containment!” shouted a worker identified as Tom. “Shut the damn thing down and quit screwing around!” Expecting a scoop on the latest leak, Nordland called the makeshift N.R.C. press center in Middletown Borough Hall for confirmation. Karl Abraham, formerly a Philadelphia Bulletin science writer and now a regional information officer for the N.R.C., told Nordland he would check it out and get back to him. In Philadelphia the presses were stopped to await the story, but Abraham didn’t make deadline. Finally he called back and told Nordland there was no leak. According to the Inquirer, which obviously had it on background, Abraham then called the governor’s press office and reported that the Inquirer had found out about a minor leak but not to worry, he had denied it completely.

At Denton’s briefing the next afternoon the Inquirer triumphantly read its radio transcript and asked for comment. Denton readily confirmed the leak, noting that small leaks were common when workers took samples of coolant water. Wednesday’s Inquirer played the story at the bottom of page one. Metro editor John Carroll says he would have done more with it but Philadelphia Orchestra conductor Eugene Ormandy retired and grabbed the headlines. “It was a small leak,” admits Carroll, “but it was the first time the N.R.C. was caught deliberately putting out false information.”

The incident has become a focus for charges of an N.R.C. cover-up. Nordland maintains that Abraham lied and that he even apologized to Inquirer reporter Ray Holton the next day for having done so. But Abraham says that he was “set up” by the paper, which was vague about how it had heard about the leak and gave him the impression that it was a major one. He did call press secretary Paul Critchlow, he says, but only to “alert him that he might get some press action.” Denton, in any case, only enhanced his standing by the way he confirmed the leak at his briefing the next afternoon.

Although it is hard to prove that the N.R.C. lied, public relations are a recurring theme in the commission transcripts. As The Washington Post summarized them on April 13, the transcripts show that the N.R.C. “worked hard to make sure that mainly ‘reassuring’ information would reach the public.” The commission wasn’t in session Saturday night when the A.P. bubble story hit the wires,but its reaction to U.P.I.’s bubble story Friday afternoon was angrily defensive. Certain that the media would grossly exaggerate anything it said about the danger, the N.R.C. consistently understated its own anxiety. The prediction was self-fulfilling, of course. Believing that the N.R.C. was grossly minimizing the threat, the media magnified the commission’s public statements on the likelihood and imminence of disaster. By so doing they achieved a roughly accurate picture of the N.R.C.’s private assessments. (At hearings in Washington in the wake of the accident, N.R.C. staff members testified that the core had been less damaged than they had thought and that the bubble could never have exploded. At the time the N.R.C. had sought unsuccessfully to hide its fears; now it said they had been excessive.)

The typical Three Mile Island story seesawed carefully between the looming threat of disaster and the industrious optimism of the experts. Imagine yourself at your breakfast table Saturday morning reading this A.P. overnight (we’ve put the bad news in roman type, the good news in italics):

Scientists struggled to cool down the stricken Three Mile Island nuclear power plant today, but authorities said the chances of a catastrophic melt-down were “very remote” and assured 130,000 nearby residents they were safe.

While technicians tried to “bleed” a bubble of radioactive vapor threatening the plant’s damaged nuclear core, Gov. Dick Thornburgh said at a news conference late yesterday that no general evacuation of the area is necessary “at this time.”

After the tensest day since Wednesday’s plant accident, Harold Denton, director of operations for the U.S. Nuclear Regulatory Commission, somewhat eased the mounting worries of local residents when he said there was “no immediate danger to the public.”

Like most T.M.I. coverage, this is fair, understandable, and accurate; concerned but calm. It is also almost mythic: paladins labor night and day to overcome the forces of chaos. Missing from the story is whether they knew how.

A Technical Tale

At the beginning, at least, the vast majority of reporters had no idea what anybody was talking about. Anchorless on a sea of rads and rems and roentgens, of core vessels and containments and cooling systems, they built their stories around the discrepancies between sources, confident that the news, when they finally came to understand it, would center on the facts in dispute. What is surprising about the T.M.I. coverage that emerged is not that it was sometimes technically wrong, but that it was so often technically right.

It’s Monday noon in the Middletown Borough Hall gymnasium, and Denton and Mattson are holding court before some 200 reporters who have battled for a precious spot where they can both see and hear. Denton stands on the free-throw line, half hidden behind a forest of microphones. There’s no PA system yet, and transcripts are still iffy, so you have to listen hard. Curious townspeople drift in and out of the bleachers, marveling at the cameras and frowning at the cuss-words. The big news is the shrinking bubble, of course, but later someone asks how the accident started. Here’s Mattson’s answer, praised by reporters as the clearest and most complete chronology of what had happened:

The steps leading to the situation we’re in today can roughly be characterized as the loss of feedwater on the secondary side of the power plant;a rise in pressure on the primary side of the reactor power plant; a discharge of coolant through the pressurizer; the initiation of the backup safety injection system, the emergency core cooling system which comes on automatically on high containment pressure, or loss of coolant from the facility; the continued high pressure of the facility with the high pressure injection system actuated; then the turning off of the safety injection system for some period of time; … and shortly thereafter the reinitiation of the emergency core cooling system after a gas situation that develops in the reactor core, for the emergency core cooling system by itself without a loss of cooling accident was unable to keep down the temperature in the core. That was finally stabilized by reestablishing the flow of the primary coolant and by restarting the main reactor coolant pump, the one that is still running.

That’s it – no wall chart, no glossary, no technical experts hanging around afterward for questions. Okay, now write a few grafs on what went wrong at the plant. Or walk across the gym to your phone, call your station, and tell your listeners how the accident happened. (When that’s done, try to discover if the N.R.C. is hiding anything.)

The reporters who were given this technical assignment fell into four categories, judging by responses to questionnaires we handed out.

  • About a third were frankly bewildered. They relaxed when the questioning turned momentarily nontechnical, took frantic verbatim notes when the going got rough, and stopped writing altogether when it got rougher still. They seldom asked questions, but the questions they did ask often brought the briefings back to basics: “Just how dangerous is this?” “What are the chances of an evacuation now?” These were the reporters whose recurring nightmare was that Denton would announce a meltdown in technical language and they wouldn’t realize what had happened. After each news conference they gathered in small knots to compare notes, and whenever possible they checked their stories against wire copy before filing. Included in this category were most of the non-network broadcasters, reporters from small newspapers, almost everyone who covered the story alone, and almost everyone who arrived later than Saturday (“fresh meat,” rotated in because of radiation anxiety).
  • Only a handful of reporters knew much about nuclear power before they reached T.M.I. They were all science and energy writers – although not all the science and energy writers qualified. At the briefings they asked highly technical questions. Between briefings they badgered the information people to call the plant for specific data, quickly collecting an audience of colleagues who wondered aloud what they wanted to know that for. Stuart Diamond of Newsday typified this breed. Most out-of-towners landed in Harrisburg with only the clothes on their backs and a notebook; Diamond brought along his Rolodex and reference books from past nuclear stories, and spent hours on a borrowed phone at the Harrisburg Patriot calling his network of expert sources.
  • Nearly a quarter of the reporters had a single expert on tap – a source from an earlier story, a science writer back in the office, or a paid consultant on the scene. These reporters often read their questions at briefings, stumbling over the jargon; they couldn’t ask follow-ups until they had checked with their expert to find out what the first answer meant. If the expert was nearby, the system worked. NBC, ABC, and the Chicago Tribune all got good mileage out of their technical consultants, but CBS probably got the most for its money. Long Island radiologist Harry Astarita arrived Friday evening to check the radiation badges of the CBS crew. Quickly dubbed “Radiation Harry,” he wound up checking copy as well – and prompting reporters with questions, scotching phony stories, and correcting false analogies. He even managed an exclusive two-hour off-the-record interview with a Met Ed engineer to get the reactor schematics straight. We wanted to talk to Astarita. He could only spare us a minute, he apologized happily: “I have to make a deadline.”
  • The rest of the reporters made themselves into experts – fast. In the wee hours of Sunday morning, reporters at Lombardo’s (Harrisburg’s plushest bar) clustered around Ben Livingood as though he were Harold Denton himself, while Livingood drew diagrams on cocktail napkins to explain the dynamics of the hydrogen bubble. Livingood is a political reporter for the Allentown Call-Chronicle. “I don’t know what happened,” mused a colleague. “None of us knew anything about this stuff Wednesday. Ben went away for six hours, and when he came back, he knew it.” On Friday, Livingood turned to a Penn State nuclear engineer for help and he continued to use him on background throughout the crisis. But, like many other reporters, Livingood himself acquired an impressive amount of nuclear expertise at T.M.I. (It helped, he says, that he “tried not to get stuck on the conflicting-sources angle.”)

The science writers on the scene weren’t a lot better off than aggressive political reporters like Livingood. “Unless you were a nuclear scientist,” ABC energy specialist Roger Peterson later told The New York Times, “you didn’t know what on earth was going on” – and science writers are seldom nuclear scientists. There were perhaps forty full-time science journalists at the site, about a fifth of the nation’s total. Some, like the Milwaukee Journal’s Paul Hayes, covered the story alone, while others, like The Boston Globe’s Jerry Ackerman, were part of a team. Either way, their stories on the breaking news were almost indistinguishable from what general assignment reporters produced. They did have a head start on the backgrounders, but many editors decided that the backgrounders could be written just as well back at the office. The big advantage went to those few science writers with a personal file of expert contacts. While Newsday’s Diamond and the few others who had covered nuclear power extensively ran up long-distance phone bills, their colleagues scrambled – often unsuccessfully – for someone knowledgeable to interview in town.

Left with the flacks

Even biased sources were scarce. Anti-nuclear experts Ernest Sternglass and George Wald came to Harrisburg for a press conference on Thursday, and then took off again. Mobilization for Survival, a Philadelphia-based anti-nuclear group, and Three Mile Island Alert, a small Harrisburg affiliate, co-sponsored the Sternglass and Wald visits, but reporters who called Alert’s unlisted phone number over the weekend were lugubriously informed by an answering service that everyone was gone and would stay gone until it was safe to return.

The nuclear industry wasn’t afraid of radiation; it was afraid of reporters. Hundreds of industry experts crowded the nearby motels, ID tags and dosimeters clipped to their lapels. The Atomic Industrial Forum, never before at a loss for words, sat vigil with Met Ed’s publicists in their suite at the posh Hershey Motor Lodge, but it gave no interviews.

Any expert at T.M.I., however biased, would have had a captive audience of hundreds. In the absence of experts, PR people ruled the roost. And charges of cover-up aside, the PR logistics were intolerably bad. Karl Abraham arrived from the N.R.C. regional office in suburban Philadelphia on Wednesday, but he had no back-up until Friday afternoon and no press facility, apart from Critchlow’s office, until Sunday night. Critchlow and his deputy, Roland Page. meanwhile, were busy asking questions on behalf of the governor; they left the question-answering to the junior staff, who mostly just took messages.

Even getting through to the N.R.C. by phone was virtually impossible during the first four days. And when a reporter did reach the office, more often than not the question had to be relayed to a technical specialist, a task that took hours more. Logistics slowly began improving when the N.R.C. got its Middletown facility in operation Monday. By then the commission had emptied its regional offices and had seven of its ten PR professionals in Pennsylvania and the other three in Washington. But they were quickly overwhelmed by the sheer volume of the requests. Moreover, until Monday night, when safety engineer Robert Bernero arrived, the N.R.C. did not have a single technical expert assigned to talk to reporters at T.M.I. Tom Elsasser, an engineer who normally handles relations with state officials, had filled the vacant briefer’s slot Sunday and had done his best.

Running with the pack

Meanwhile, reporters helped each other. Viewed from the desk, T.M.I. was a hotly competitive story, and few reporters escaped a daily call from an editor asking why they had missed an angle. But viewed from the site, the story required collaboration. From the moment the Harrisburg press corps heard about the accident, explains The Philadelphia Inquirer’s Tom Ferrick, “we all shared information. We got drawings and pieced together the sequence of events at the plant. We went out and got books on nuclear energy and compared them and discussed how a reactor works.” Newspaper Enterprise Association columnist Bob Walters calls this “pack journalism at its worst.” To reporters trying to get a technical story right, without ready access to the experts, it seemed like good sense.

Three Mile Island offered few alternatives to pack journalism. Reporters occasionally found a fresh color angle, or broke away from the herd with a background piece – but the main event was inside the T.M.I. reactor, and information on that event had to come from the N.R.C. or Met Ed. A few individual news organizations managed to stand out – the A.P. with its detailed play-by-plays of conflicting statements; The Washington Post with its unmatched contacts at N.R.C. headquarters (the Post covered the story more from Washington than from Pennsylvania); The Philadelphia Inquirer with its all-out muckraking zeal. But enterprise reporting rarely dominates a breaking story, certainly not a breaking technical story. Most hard news pieces on T.M.I. were interchangeable. And most of the reporters who wrote them collaborated like unprepared students on an unfair homework assignment: “What have you got for how big the bubble is this morning?”

When they got something wrong, they all got it wrong. Most reporters parroted misleading information about radiation exposure for days – even weeks – after the accident. At the beginning they ignored the differences between the radiation dose per hour in a plume that passes by, the cumulative dose received by a person who stays in Middletown, and the continuing radiation from a particle of Iodine-131 that is absorbed by the thyroid. Some of these errors were cleared up by the end of the week, but Met Ed’s false comparison of radiation-leak dosage with X-rays lasted much longer: your whole body is not exposed to a chest or dental X-ray, a difference reporters unconsciously acknowledged by crossing their legs whenever radioactivity was discussed.

The news about radiation at Three Mile Island was generally reassuring. But somehow neither readers nor reporters were reassured. On April 9 Bill Drummond did a sensitive piece on radiation anxiety in the Los Angeles Times, based largely on interviews with psychologists. He could as easily have interviewed reporters. “I felt safer in Biafra than I do here,” said Ray Coffey of the Chicago Tribune. “I’d be a lot happier if we could paint this shit purple or make it smell.”

Milking the Story

By Monday afternoon the crisis was waning, and so were the coverage problems. Reporters were beginning to understand the technical details; the N.R.C. was beginning to get its logistics together; the conflicting accounts were beginning to agree. Jeffrey Hodes of the Independent Television News Association was nonetheless depressed. “I feel like a failure,” he confessed. “I cannot understand this stuff and it’s impossible to explain it in sixty seconds. I finally cornered a technical guy from the N.R.C. and said, ‘Look, I’ve been here four days and there’s just one thing I want to know: exactly what’s wrong.’ He started to explain it to me. He drew a diagram and told me how the pressure built up like when you shake a pop bottle. For the first time I was beginning to get an idea of what was going on in the reactor. Then Jimmy Breslin walked up and asked him, ‘When are you going to release da names of da woikers? [Four Met Ed employees had exceeded their permissable radiation doses.] Huh? What about da woikers?’ Six other reporters rushed over to hear the answer, and that was the end of my lesson.”

Like Breslin of the New York Daily News, about half the reporters who haunted Harrisburg in the days after the accident studiously avoided what they called the “plumbing” story, relying for that on colleagues or the wires. They were there to cover the conventional disaster story that never materialized. While they waited, they covered the preparations for evacuation. And they wrote color pieces.

Color was more than a legitimate part of the T.M.I. story; it was an important part. Poll results since the accident show that fear is a crucial new element in the energy cost-benefit calculation: reporting this fear – or its absence – is significant journalism. That said, we find much to complain about in the color coverage at Three Mile Island.

Evil steam

Editors (especially on TV) sometimes let the sidebars overwhelm the story, leaving the impression that fear was the main event at T.M.I. In the week after the accident, for example, ABC used thirty-nine interviews related to T.M.I. on its evening news. Nine were with Denton, Hendrie, or other N.R.C. sources; nine were with local, state, or national politicians; three were with Met Ed representatives; one each was with a plant worker, an insurance adjuster, and a radiation expert. The other fifteen were with local residents, who told how scared they were, or weren’t. Although ABC’s reporters at the site did a competent job of summarizing technical developments, people who followed Three Mile Island on this network saw more of frightened farmers and brave children than they did of Harold Denton and his aides.

More important, the color coverage was one-sided and unsophisticated in its handling of scientific controversies. Newspapers and networks regularly crossed the line between covering people’s fears (legitimate and important) and letting those fears represent the actual situation (misleading and unfair). The words of a dairy farmer who claims he’s afraid to drink his own animals’ milk more than balance official statements that the milk is safe – especially when the farmer is positioned to speak for both sides. Here is Jim Hardison, interviewed by Bill Zimmerman on ABC after a state inspector checked his milk: “He said there’s no iodine showing up and at this point it’s perfectly, well you know, to drink. But I won’t take that chance. … I don’t think we’ve been informed in the past ten years the amount of radiation that’s been released from that plant.” Such claims – some of them hotly debated, some discounted even by anti-nuclear experts – needed to be evaluated, or at least balanced in the same story.

Inevitably, perhaps, the color coverage was often superficial and stereotyped. The kids were all cute or handicapped (or both); the farmers were all struggling; the senior citizens were all feisty. Sometimes the misrepresentation was more serious. “I’ll give you just one example,” said an exhausted Red Cross volunteer two weeks after the accident. “I’ve been doing radiation screening, to reassure people. Today some reporters from Washington asked me to help them find a pitiful case to interview. That’s what they said, a pitiful case.” Earlier, Atlanta Constitution reporter Barry King had watched a network crew on the streets of Middletown asking people to please “stay out of the camera shot.” The residents obliged, King wrote, but “they were later angered to see their town depicted as an ‘abandoned city.’”

There was nothing to see at Three Mile Island. Despite Breslin’s Daily News rhetoric about “evil” steam dripping down the cooling towers “like candle wax,” T.M.I. at the height of the crisis looked exactly like T.M.I. before the crisis – except that the two-lane road across the Susquehanna from the plant was crowded with broadcasters taping their standuppers. A proposal to put a pool camera in the control room got nowhere. The most frustrated journalists at T.M.I. were the photographers and TV people (Life later astounded everyone by managing to make the story visually interesting). Their editors clamoring for fresh front-page fodder, the word people went looking for someone to interview. With sources hard to find and readers already overdosed on technical detail, they wound up writing color. Says Peter Stoler of Time: “You want a panicky citizen, you get a panicky citizen.”

A glimpse of Armageddon

The color at T.M.I. was overcovered, but T.M.I. itself was not – despite charges of sensationalism from the White House and elsewhere. Every overblown headline like the New York Post’s RACE WITH NUCLEAR DISASTER had its understated counterpart like the Manchester Union Leader’s NO INJURIES REPORTED IN NUKE MISHAP. Walter Cronkite couldn’t resist glimpsing Armageddon on the horizon – but all-news WCBS radio in New York stuck stolidly to its format, seldom according T.M.I. more than sixty seconds at a shot. Most of the coverage stayed responsibly in the middle.

Nuclear proponents and opponents alike were bewildered by the T.M.I. media blitz. The system worked, industry supporters repeated, just as it has worked in other nuclear accidents: there was no catastrophic meltdown, no big explosion, not a single fatality. Said anti-nuclear activists: it’s what we’ve been telling you all along. Where were you at all those other accidents?

What, indeed, made T.M.I. such a story? For one thing, the accident came at a time of heightened awareness of nuclear issues. The China Syndrome was proving a box-office hit and had already made “meltdown” a household word. PBS had just aired “Paul Jacobs and the Nuclear Gang,” a strong statement about radiation danger and government indifference. The N.R.C. had recently found flaws in the Rasmussen Report, which held that meltdown was extremely unlikely, and had then closed five northeastern nuclear plants because they were no longer judged to be earthquake-proof. Congress was considering a bill to expedite plant licensing, and the commission was considering a cut in radiation exposure limits. The place was right, too. Harrisburg is just a short hop from New York and Washington, with plenty of hotels, restaurants, and plane connections.

The accident itself was unique. The rich smorgasbord of conflicting statements served up by Met Ed and the N.R.C. piqued editors’ interest in a possible cover-up. And they had time to go find out. Most nuclear accidents, however dangerous, are over in hours. This time the cliffhanger lasted for days.

Most important, Three Mile Island marked the largest release of uncontrolled radiation into a heavily populated environment in the history of the industry. The radiation triggered a “general emergency” that forced plant authorities to warn the public promptly – and then the radiation kept on coming. In the week following the accident, a hypothetical resident standing naked at the T.M.I. north gate absorbed roughly eighty-five millirems of radioactivity, less than the annual difference between what residents of Harrisburg and Denver receive. (Reporters, according to the dosimeters many brought with them, absorbed well under ten millirems.) The resulting increment in the local cancer rate, though arguable, will certainly be small. But reporters didn’t know that Wednesday afternoon when Lieutenant Governor Scranton first announced that there had been two puffs of radiation. They didn’t know it Friday morning when officials admitted to a third substantial release. They knew only that radiation was leaking at Three Mile Island – sporadically, invisibly, and dangerously.

Reporters came to T.M.I. to cover two stories, the leaking radiation and the conflicting statements. They were soon engulfed by a bigger story, the battle against the bubble. Despite the obstacles – the scarcity of sources, the rotten logistics, their own fear and ignorance – they handled all three stories reasonably well. And as the crisis subsided, they finally began to focus on the fourth and most important story: Do the experts know enough to protect us from nuclear catastrophe? That story has been around, largely uncovered, for a decade. Now it is news.

Copyright © 1979 by Columbia Journalism Review

Sidebar articles:

For a 2004 retrospective on the Three Mile Island accident, focusing on its enduring lessons of crisis communication, see “Three Mile Island – 25 Years Later,” ( 106 kB) on this site.

For more on crisis communication:    link to Crisis Communication index

      Comment or Ask      Read the comments
Contact information page:    Peter M. Sandman

Website design and management provided by SnowTao Editing Services.