Andrew Hale, Delft University of Technology
Safety of aviation – most likely a political compromise
An argument that the binding constraint on aviation safety is governance, not instrumentation.
Andrew Hale’s essay carries the most sceptical title in the project — safety of aviation, “most likely a political compromise” — and it earns it. His argument is that safety is not a technical trajectory that improves as instruments improve, and that anyone projecting 2093 from hardware is looking at the wrong variable.
The opening definition does the work. The safety level a system reaches is the result of removing evident causes of failure in technology, human factors, interfaces and management, but also “of compromises in which safety is traded off against other system gains, such as airspace capacity, operation in extreme weather conditions, cost and political pressures”. On that basis the level reached in 2093 “will be probably more dependent on how the pressures affecting those compromises are regulated and managed, than on the changes in hardware and instrumentation”.
The two gaps he identifies
Hale is writing as a professor of safety science at Delft, and he does not soften the geography. In 2008 there was a clear difference in safety between regions and between the airlines based in them: Europe, North America and Australia averaging safer than many parts of South America, Africa and Asia.
The second gap he considers larger. Between commercial airline travel and the rest of aviation — freight, business, private flying, and what he calls marginal forms such as microlights and hot air balloons — the differences are “equally large, if not larger”. Some of that is genuinely technical: aircraft age, airport instrumentation, weather, terrain. But he refuses to leave it there. The differences are also “due to the enforcement and acceptance of lower safety standards” in those regions and sectors.
That distinction between capability and acceptance is the essay’s pivot. A standard that exists but is not enforced produces the same accident rate as no standard.
The maritime comparison
His analogy is flags of convenience: the practice of registering ships in countries with lax standards and weak enforcement. Shipping’s answer was port state control, which lets a country inspect visiting vessels in its ports and detain them until they are made safe, or refuse entry altogether.
He rates the result honestly: mixed success, “because it is not always easy to determine from a physical inspection whether there is danger present.” A hull can look sound and be badly run.
Aviation, he notes, had begun moving the same way. The United States and the European Union operated blacklists of airlines forbidden to land, based on inspections and safety record, and other countries had followed. This is where his political argument bites. Landing rights are reciprocal currency: a request for a European country’s aircraft to land in an African country “may only be honoured if that country’s planes are allowed to land in the European country; yet another compromise.” Safety enforcement and market access are entangled, so a regulator acting on safety is also acting on trade.
His actual recommendation
The essay ends with a claim about where effort should go, and it is unfashionable. The biggest available improvement by 2093 comes from “ensuring that all countries and aviation sectors adopt the technology, practices and safety standards of the current best, rather than trying to make the ‘ultra safe’ ultra, ultra safe.”
In other words: stop optimising the already-excellent top of the distribution and raise the bottom. Marginal risk reduction on a Western scheduled airline is expensive and buys little. Bringing a poorly regulated sector up to existing best practice buys a great deal, using technology that already exists.
Then the question he leaves open, and does not answer: “whether there is the political will and societal pressure to do so?”
How it reads now
Better than most of the project. Regional divergence in aviation safety remained a live problem long after 2008, and the EU list of banned carriers stayed in force and in active use. The distribution argument also held: the large accident totals of the following years clustered in exactly the sectors and regions Hale pointed at, not at the safest end he advised against polishing further.
The essay’s weakness is what it declines to do. Hale never estimates how much of the gap would close, or on what timescale, and his final question is left as a question. For an essay commissioned to look eighty-five years ahead, it commits to remarkably little — which is either its discipline or its evasion, depending on how much you trust forecasts.