At about sixteen thousand feet above Oregon, six minutes after leaving Portland on the evening of 5 January, a rectangular panel in the side of a Boeing 737 MAX 9 operated by Alaska Airlines left the aircraft. The cabin decompressed. The crew turned back and landed at Portland. Nobody was seriously hurt, which had a great deal to do with the two seats beside the hole being empty.
Within a day the Federal Aviation Administration had grounded the type. Twelve days later the grounding is still in force and the cause is still not established. This is a primer on what happened between those two sentences, because the machinery of grounding an aircraft is reported constantly and explained almost never.
What the panel was
The object that departed is called a door plug. Some airlines configure the 737 MAX 9 with enough seats to require an additional emergency exit; others do not. Rather than build two different fuselages, the manufacturer builds one with the opening in it and fits a permanent plug where the door would be on the higher density version.
A plug is not a door. It does not open, has no handle, and is bolted into place to become part of the pressure vessel. From inside the cabin it is invisible, sitting behind a normal wall panel and a window.
So a passenger in that row is sitting beside a structural component that is not supposed to be removable and that a maintenance organization can, in principle, remove.
What a grounding actually is
The FAA did not telephone airlines and ask them to stop. It issued an Emergency Airworthiness Directive, which is a legal instrument with a specific and narrow character.
An Airworthiness Directive is a rule. It applies to a defined set of aircraft by type and serial number, states an unsafe condition, and specifies what must be done before further flight. Operating an aircraft covered by an active directive without complying is not a policy breach; it is a violation of federal regulation.
The emergency version skips the ordinary notice and comment period, because the unsafe condition is present now. That is the whole of its power: it can be issued in hours and it binds immediately.
What it does not do is say the aircraft is unsafe in general terms. The directive names a condition, and lifting it requires a defined inspection, not a judgment that enough time has passed.
Who is investigating, and who is not
Three organizations are involved and they are frequently confused, which matters because they have different jobs and different powers.
The National Transportation Safety Board investigates. It determines what happened and why, issues recommendations, and has no regulatory authority whatsoever. It cannot ground anything, fine anybody, or compel a design change. Its power is entirely in the quality and publicity of its findings.
The FAA regulates. It certifies aircraft, oversees manufacturers and airlines, and issues the directives. It can ground a fleet and it can compel modifications.
The manufacturer builds, and holds the type certificate, a relationship our Business desk has followed since the first MAX grounding. Under longstanding arrangements it also performs a substantial share of the certification work on the regulator’s behalf, through designated employees, which is an efficiency and a structural tension that has been argued about for a decade.
The separation exists for a reason. An investigator who could also punish would find people less willing to talk, and safety investigation depends almost entirely on people talking. So the body that learns what happened is deliberately denied the power to act on it, and the body with the power depends on the other one’s findings.
Why the inspections take so long
An inspection order sounds like a checklist and is not.
Before an airline can inspect, somebody has to write the procedure: which panels come off, what is examined, what torque values are expected, what constitutes a pass, what is done about a fail. That document has to be produced by the manufacturer and accepted by the regulator, and until it exists the aircraft sit.
Then the work is done by licensed engineers, of whom there is a finite number, on aircraft that are in whatever locations the schedule left them in when the directive landed. An airline with sixty five affected aircraft does not inspect sixty five aircraft simultaneously. It inspects as many as it has qualified people and hangar space for.
And each inspection can generate findings that were not the subject of the directive, which then have to be resolved before that individual aircraft flies.
What it costs, and who carries it
A grounded aircraft costs money continuously. Lease or ownership costs do not pause. Crews are paid. Slots at airports are held or lost. Passengers are reaccommodated onto other carriers at whatever the market rate is that day, or refunded.
Those costs fall first on the airlines, which did not build the aircraft. Some of it is recoverable from the manufacturer later, through compensation negotiated privately and rarely disclosed, but the cash goes out now and comes back, if it comes back, in a future quarter.
The second cost is harder to see and larger. A regulator that has grounded a type has made a public statement that it did not previously know something important about an aircraft it certified. Every subsequent approval that regulator issues is read in that light, by other regulators who have historically accepted its certifications without repeating the work.
There is a further cost that lands on nobody in particular and is real anyway. Every crew, dispatcher and engineer working the type now carries a question they did not have a fortnight ago, about a component they had no reason to think about. That does not show up in a filing, and it is the part that takes longest to put back.
The part nobody can resolve yet
Twelve days in, the honest position is that the cause is not publicly established. Investigators have the plug, recovered from a garden in Portland. They have the aircraft, the records, and the people. They do not yet have a published finding, and they should not be expected to: an investigation that produces an answer in a fortnight is usually producing a guess.
Which leaves everybody in the uncomfortable middle. The fleet is grounded against a condition that has been described but not explained. The inspections will find whatever they find on individual aircraft, and a clean inspection tells you that one aircraft is correct now, not why another one was not.
The plug is in a laboratory now, and whatever is learned from it will arrive months from now, in a document written by an agency that cannot order anybody to do anything. The agency that can has already acted, on the strength of not knowing. That is the system working as designed, and it is the hardest part of it to explain to somebody holding a boarding pass for Tuesday.
