The Engineering Deficit.

Armchair generals are going on industrial chemistry courses.

There has been a retraining programme under way, with the armchair general picking up a second discipline. It used to be maps and motives, arrows and choke points and the question of who blinks first. Now the same person is half way through a crash course in industrial chemistry, trying to work out why sulphur counts, what helium actually does and how something called naphtha ends up in everything from packaging to paint, because the war has not changed character so much as the relevant inputs have come into focus.

What this exposes, almost by accident, is a gap in how people are trained to think about the world. The dominant voices move easily through politics, narrative and emotion. They can describe intent, speculate on motives and shape a sequence of events into something coherent, which carries real value until the question shifts from why something is happening to how it actually works. At that point chemistry and engineering take over, because they deal in systems that do not respond to tone, persuasion or protest. A reaction proceeds under defined conditions or it does not, a plant runs within tolerances or it stops. When an input goes missing the chain breaks in a manner no statement can repair.

The people describing that failure were never trained to see it coming. Engineering graduates tend to leave. They move into finance, consulting and industry, where the problems are concrete and feedback is precise. The people who remain in public commentary are those whose skills suit commentary, language, synthesis and persuasion. The standard route into journalism runs through politics, English, history and classics. Science graduates who enter the profession tend to find their way to specialist publications, leaving the energy and infrastructure brief to people whose training optimised for narrative rather than process. No individual failing, it is only a selection effect.

Engineering training forces you to hold two things together at once. The hard boundary of physics and chemistry on one side, the creative act of designing within it on the other. You are not merely learning facts but learning how to make something function inside limits that do not move. One of those limits is tolerance. The question is never whether something is perfect but whether it is within specification and tolerances compound. A component within tolerance combined with another within tolerance combined with a third can still produce an assembly outside tolerance, which is why engineers think about degraded conditions, load and deadline rather than judging proposals against the ideal. The gap between design specification and operating reality is where most things fail. Engineering training is specifically about that gap and almost nothing else is.

It also enforces a habit of mind. A design either works or it does not. A process either runs within specification or it stops. The question of what would have to be true for this to fail becomes second nature and that habit transfers into any field where outcomes are testable. Commentators struggle with this because the feedback loop is long and diffuse. A wrong prediction can be reframed, a structural failure cannot. The engineer who specified the wrong beam does not get to recontextualise the collapse.

Engineers design for failure modes, not only for function. Failover, single points of failure and slack are standard concepts in any systems discipline and invisible in policy design. Infrastructure built without redundancy looks efficient until one component fails, at which point the efficiency was never real. Several decades of just-in-time optimisation removed the slack that made supply chains robust, which looked like a productivity gain until it became a fragility and the inputs that stopped arriving have not come back.

The department currently running Britain’s net zero programme, its most physically demanding peacetime commitment, demonstrates the sorting argument. The Secretary of State holds a degree in Philosophy, Politics and Economics from Oxford and a master’s in economics. The Minister for Energy read history and politics before retraining as a teacher. The Minister of State took a doctorate in political science. Not one person across the ministerial layer, the senior civil service or the advisory board holds a background in science, engineering or mathematics. The head of Mission Control for Clean Power 2030 studied finance and law. The chief executive tipped to lead the Climate Change Committee read classics at Oxford and began her career as a climate campaigner.

The name Mission Control borrows from one of the more famous engineering operations in history, where rooms full of aerospace engineers read telemetry in real time and understood immediately what it meant. The UK version is described officially as a one-stop shop to troubleshoot, negotiate and clear the way for energy projects. Troubleshoot is an engineering verb. Negotiate is a political one. The team assembled has the negotiating background and when the grid fails to balance that difference will show up somewhere other than a press release.

When production moves offshore the engineering workforce that maintained it either retrains, retires or follows the work. The UK has been running this experiment since the 1980s and net zero accelerates it. Carbon accounting that treats imported steel, cement and chemicals as invisible in the domestic accounts makes it rational to close production here and buy it finished from elsewhere. The knowledge does not transfer with the invoice. It retires with the people who held it and the next generation does not replace them because the work that would have trained them no longer exists.

The finished goods arrive and nobody has to think about how they were made. A country that makes things knows what making things requires, the tolerances, the sequencing and the points where the process breaks. A country that buys things finished can persuade itself that the making is a detail someone else handles and eventually that persuasion becomes the default assumption in the institutions that set policy. Engineering does not disappear from the world. It relocates to South Korea, Germany, Denmark, Finland and becomes, as far as the buying country is concerned, a logistical question rather than a technical one.

People stop choosing engineering because the economy appears not to need it. If the engineering is happening elsewhere and the goods arrive on time, fewer people see the point of the subject at home. Fewer students choose it, fewer engineers enter public life and fewer arrive in policy roles and the people designing industrial strategy have no personal experience of what production actually involves.

Net zero then lands on this foundation. The policy most dependent on engineering capability is being designed by people with no engineering background, for a country that has spent four decades exporting the engineering base the policy requires. The offshore wind turbine components come from Denmark and Germany. The grid cables come from Finland. The battery chemistry comes from Asia. The planning system that blocked onshore wind also blocked the industrial expansion that would have kept those skills onshore. Each decision looked reasonable on its own. Together they removed the floor the current programme is trying to build from, on a fixed deadline, designed by people who never had to ask where the floor went.

PPE and political science produce sharp analytical minds, well suited to persuasion, coalition building and managing competing interests, which is the preparation needed to win the job. Winning the job and doing the job are different tasks when the job involves specifying grid capacity, sequencing offshore wind connections, understanding what grid balancing costs under load, or evaluating whether a 2030 clean power target is physically achievable within the infrastructure that will exist by then.

A systems programme fails among the people running it in one of two ways. The first is not understanding the physical limits, which produces targets and timelines that were never achievable. The second is understanding them and overriding them when political pressure demands it. Angela Merkel held a doctorate in quantum chemistry and still closed Germany’s nuclear plants after Fukushima, following the political wave rather than the physics, with the result that Germany burned more coal, missed its own emissions targets for years and paid more for power throughout. Technical training removed neither the political incentive nor the willingness to act on it. What its absence does is add to the exposure that is already there, because the minister who cannot evaluate technical advice cannot identify the moment when political pressure is asking them to override a hard limit. Merkel at least knew what she was overriding.

Singapore chose differently, at least for a period. Lee Hsien Loong, who governed from 2004 to 2024, was Senior Wrangler at Cambridge, graduating with first class honours in mathematics and a diploma in computer science. The training produces a specific habit of working from first principles, tracing dependencies, asking what has to be true rather than what would be convenient. That carries no guarantee of good governance. When advisers told him something was impossible within a given budget or timeline, he had the tools to evaluate whether they were right. The minister who cannot read the technical argument cannot distinguish between an adviser being honest and one telling them what they want to hear. China ran a deliberate technocratic programme from the 1980s through the turn of the century, when the majority of its ministers, provincial governors and party secretaries held engineering degrees and what got built in that period reflected it. Legal and economic training gradually displaced engineering in the leadership pipeline. The instinct for asking what the system could actually bear went with it.

Markets run on stories, but the real work sits underneath in flows, positioning, liquidity and collateral. An engineering mind looks at a price and asks what sustains it, what funding structure supports it and what happens when one leg is removed. Politics and government planning follow the same logic. Energy systems, transport, housing, water and defence all resolve into questions of capacity, throughput and failure points. Those questions carry hard boundaries whether ministers acknowledge them or not. Housing requires materials, labour and infrastructure in sequence. Electrification requires generation and grid capacity before it requires anything else. Transport upgrades require spare capacity and some respect for bottlenecks.

Naphtha feeds into packaging, packaging into cold chain logistics, cold chain into food security, food security into social stability. Five links and the commentariat habitually engages with the last one while remaining incurious about the first four.

The industrial chemistry course is not a hobby. It is the recognition that the causal chain runs through inputs nobody was previously watching and that following it requires a different set of tools than the ones the commentary class spent years sharpening. The armchair general now follows sulphur down the chain. Sulphuric acid to fertiliser, mining, oil refining, chemicals, batteries, agriculture, food production, autos, construction, electronics, pharma, textiles, pulp and paper, water treatment, energy systems, transport and defence.