The Window Closes Before the Aircraft Arrives

Aviation is the part of this work that photographs well, and the photograph is misleading. An extraction reads, from outside, as the arrival of an aircraft — as though the scarce thing were lift, and the problem a procurement problem solved by having the number of somebody who has aeroplanes.

Lift is a commodity. On any ordinary day a broker can find an airframe and a crew inside an hour, and the market for that is competitive, transparent and largely interchangeable. The constraint sits somewhere else entirely, and it is not purchasable at the moment it is needed.

What the paper says, and what it does not do

Start with the law, because the law makes the point better than any anecdote could. Article 5 of the Convention on International Civil Aviation provides that aircraft not engaged in scheduled international air services have the right to make flights into or in transit non-stop across the territory of a contracting state, and to make stops for non-traffic purposes, without the necessity of obtaining prior permission — subject to the right of the state flown over to require landing. Scheduled services get the opposite treatment: under Article 6 no scheduled international service may be operated over or into a state's territory except with that state's special permission.

Read that quickly and a private evacuation flight appears to be the easy case. It is non-scheduled. The treaty says no prior permission is needed.

Now read the qualifications, because they are where the aircraft actually sits. The right is subject to observance of the Convention's terms, to the state's right to require landing, and to conditions, regulations and limitations that states are entitled to impose. In practice, that means diplomatic clearance numbers, overflight permits from each authority on the route, landing permission at the destination, and a set of national lead times that are published as working days and are not always observed as such. The right exists. It is exercisable through offices, and offices are staffed by people, and on the night in question the question is not whether the right exists but whether the person who issues the number is reachable and willing to work.

That gap — between a right on paper and a permission in hand — is the whole subject.

The dependency chain

An orderly departure is not one transaction. It is a chain, and the chain fails at its weakest link rather than its most expensive one.

There is airspace: clearance to cross each jurisdiction on the route, which is a separate application to a separate authority, each with its own hours and its own view of the requester.

There is the field itself. Under the European slot regime, a slot is the scheduled time of arrival or departure allocated to an aircraft movement on a specific date at a coordinated airport, and at a fully coordinated airport an operator needs one in order to land or take off at all. The rules do provide for this: the coordinator is required at all times to endeavour to accommodate ad hoc slot requests for any type of aviation, including general aviation, drawing on the pool and on slots released at short notice. Which is to say the accommodation is an endeavour rather than an entitlement, and it is made at a field that, in the circumstances where any of this matters, is not experiencing a quiet week.

There is ground handling, which must still be operating. There is fuel, which must still be available and payable. There is the movement from where the principal is to where the aircraft is, which is frequently the segment with the least margin and the least aviation content. And there is the documentation position of every passenger — not the principal alone but each person expected to board, which is where a departure most often loses people it intended to carry.

Each link is somebody's decision. None is a purchase.

The sequencing compounds the problem. These dependencies are not parallel, and several cannot be started until an earlier one has resolved: a slot request is made against a route, a route is built from the clearances actually granted rather than the ones applied for, and a change to any leg re-opens the ones behind it. A chain of six independent permissions, each individually probable, is not itself probable, and the arithmetic gets worse precisely as conditions deteriorate — which is when the number of parties willing to answer an unfamiliar request also falls. This is why the useful preparation is not a plan but a tested one. A plan establishes that a route exists. A test establishes that the handler answers, that the operator will accept the risk on the day rather than in principle, and that the permission which is theoretically available on a Sunday is available on a Sunday.

The relationship is the asset

In an evacuation the most valuable asset is not the aircraft. It is the relationship with the charter operator, the ground handler, the local authority and the airspace manager — assembled in advance, tested under pressure, and available to activate at the moment the decision is made. By the time those relationships are being built from scratch, the window for orderly extraction has frequently closed.

Relationships of that kind are not contracts and they cannot be manufactured under pressure. What they consist of is a handler who recognises the name, an operator who has flown for the house before and will take the call at three in the morning, an authority that has dealt with the requester on routine matters and has no reason to treat this one as suspicious. Every element of that is a deposit made in ordinary conditions and drawn down in extraordinary ones, and none of it can be built during the event.

The cost, stated plainly

Holding aviation as standing infrastructure rather than brokering it at the point of need is expensive, and most of the time it sits idle. That is not a defect of the model. It is the model. Idle capacity is what capacity looks like before it is used, and an arrangement that costs nothing while nothing is happening is an arrangement that will not be there when something is.

The alternative is honest too, and worth stating rather than caricaturing: brokered lift is cheaper, and on an ordinary movement in a functioning jurisdiction it produces the same aircraft with the same crew. The difference does not appear on the routine flight. It appears on the one where the field is congested, the handler has gone home, and the permit is being asked for outside published hours.

The sovereign who cannot move cannot decide, and that is the reason any of this is held rather than hired. A principal who is not confident of movement makes different decisions from one who is — earlier, more cautious, more concessive — and the difference shows up in outcomes that have nothing to do with aviation. Movement is not the service. It is the precondition for having a choice, and it has to exist before the moment when the choice is required.