• News
  • Guesstimation


Accident investigation is in danger of getting a bad name. The three US authorities - the National Transportation Safety Board (NTSB), the Federal Bureau of Investigation (FBI) and the Federal Aviation Administration (FAA) - associated with the EgyptAir flight 990 investigation deserve some understanding given the ferocity of the media search for information. But letting the media influence the pace and early direction of an investigation, which is what has been happening, is just unprofessional.

On 17 November NTSB chairman Jim Hall issued a press release entitled: Factual Update on EgyptAir Investigation. Hall's use of the word "factual" is an indicator of just how far into the realm offiction the investigation was in danger of being pushed. Most ordinary investigators in the NTSB were, no doubt, managing to retain a scientific approach to the real inquiry, but that is not how their attempts were being represented by Hall.

It would be easy just to blame the media for rampant speculation, but in this case the degree of misrepresentation and the apparent lack of care about the human repercussions is largely the fault of the NTSB and the FBI.

Both agencies regularly face media pressure, so have no excuse for ignorance of how the media will react to information. The media will speculate, that is definite. That they will put pressure on investigators to speculate is also certain. Intelligent speculation is a natural activity for human beings, and there is nothing reprehensible about it in its own right. For government investigative agencies to supply the media with speculation, however, is unprofessional in the extreme. Anything the agencies say carries authority; speculation can cause immense harm.

We have argued before that it is unreasonable, given the long time that it takes for accident reports to be published, for investigative agencies to withhold information when they are certain of it. A flight data recorder readout, for example, can enable details of a flight profile, and possibly of equipment failure, to be known within minutes, and that information should be put in the public domain. If it is not clear initially why the profile developed as it did, or why the equipment failed, that is a crucial part of the information that the investigators must supply. But they must not be tempted into speculation.

As for human factors, especially complex ones, it takes longer than a week to be sure of how these affect an accident. In many countries, the law dictates that voice transcripts from cockpit voice recorders (CVR) should not be published - ever.

An explanation of operationally relevant pilot exchanges is published with the report, and sometimes interim reports will paraphrase the pilot's operational exchanges. That is one way of protecting pilots against the speculation in which the media is bound to indulge if handed a literal transcript while little is known about the circumstances on the flightdeck.

Protection of the pilots, who may have had their mother tongue loosely translated and whose personal verbal shorthand or use of slang may not be understood initially, even by the investigators, is important. If pilots have acted reprehensibly, that will eventually become clear. If they have acted illegally, the same is true. But if the latter, they should face judicial procedures, not trial by the media. Especially media who have been supplied with speculation about pilot motives by the investigating agencies. In the EgyptAir case, FBI officials leaked that pilot suicide was suspected, even before the CVR was recovered from the sea. The NTSB's Hall did not deny the rumours, but said on 16 November that a decision on whether to transfer the investigation to the FBI had been delayed until Egyptian expertise had been brought in.

Whether the cause of the crash turns out to be pilot suicide or not is hardly the issue. The antics of the NTSB and the FBI have given agencies around the world a good argument for not releasing interim reports during an investigation. Some will take it as justification for not releasing accident reports at all and that's bad for the industry and bad for public confidence.