Reversals Without Reckoning
The Iran War is producing policy reversals that should have occurred years ago. Who pays the cost of delay?
The Reversals
The Iran disruption has produced something worth noting beyond the oil price and the shipping data. A series of political reversals, delivered by the people who made the original decisions, with no apparent awareness of the irony.
John Swinney, leading the Scottish Nationalist Party, a party whose energy policy since COP26 was built on a presumption against offshore drilling, complete with Nicola Sturgeon posing for selfies with Greta Thunberg, has now suggested that domestic gas carbon intensity should form part of the consideration for new North Sea licensing. He said this weeks before a Scottish election.
Grangemouth, Scotland’s last petrochemical refinery, is closing this year. Hundreds of jobs gone with it, along with the plant that used to process Forties Field oil into fuel burned at Scottish pumps, a facility that will not come back.
Ursula von der Leyen stood at a nuclear summit in Paris on 10 March and told forty countries that Europe had made a strategic mistake in turning its back on nuclear power. She noted it provided a third of European electricity in 1990 and provides fifteen percent today. Von der Leyen was a minister in the German cabinet that accelerated the nuclear phase-out after Fukushima. She then became Commission President. Her career advanced through the mistake.
The person who deepened the dependency is now positioned as the author of the relief. No process within the system distinguishes between those two facts.
What connects these reversals is the absence of cost. The argument for nuclear was always available. The case for domestic gas licensing was always makeable. What changed was not the evidence but the political price of acting on it. And when that price fell because a war moved energy markets, because an election approached, the reversal came.
This is the central problem of political accountability. The people who make consequential decisions under uncertainty face almost none of the consequences when those decisions prove wrong. The loss is socialised across an economy, the career continues. In any other system that handles decisions under uncertainty, this would be considered a design flaw serious enough to dismantle the architecture. In government it is considered normal.
The question is what a system looks like that changes this. Through structures that make the cost of being wrong land on the person who was wrong, before the position has already closed someone else’s refinery or blown up their nuclear power station.
One Dial
GDP hits first when people reach for a target, the headline metric that could in theory govern pay, tenure or consequence. It is the wrong target. A lagging aggregate that embeds damage before it moves against you, it measures activity rather than the conditions that sustain it and cannot distinguish between growth built on genuine capital formation and growth borrowed from future capacity through stimulus or asset inflation. A minister who inflated a headline number while hollowing out the underlying system would look fine until the system failed.
Goodhart’s Law operating at the level of national policy produces exactly this: once the measure becomes the target it stops tracking what it was meant to track. The desk looks fine. The P&L is lying.
Single measures invite single-point gaming. The task is to understand what a system looks like that makes gaming harder than delivery.
The Trading Desk
Every institution that handles consequential decisions under uncertainty eventually develops an error correction system. The trading desk is the clearest surviving example.
A trader is wrong more often than right. This is the explicit assumption the system is built around. The edge is asymmetry. When a position moves against you, the stop loss closes it quickly and the loss is small and defined. When a position moves with you, you hold and let it run and the gain is large and open-ended. Small frequent losses, large infrequent gains. Being right forty percent of the time produces a strongly positive outcome if the wins are three times the size of the losses. The architecture is designed around this ratio: ensure losing positions cannot overwhelm winning ones and the math works in your favour even when your judgment more often fails than not.
The entry decision is the declared thesis, explicit and falsifiable. Position monitoring runs continuously against it. If the backdrop has shifted, if price action is denying rather than confirming the view, the position is flagged. The stop loss removes the sunk cost trap entirely. When the position moves beyond a defined threshold it closes on a pre-committed rule. The embarrassment attaches to breaching the stop. The P&L records what happened across all positions, current and unambiguous, with nothing smoothed. The risk committee sits outside the desk with authority the trader cannot override. Its survival depends on outcomes.
The culture of fast cutting is the output. Changing your mind is the job. The trader who sees the thesis break, cuts immediately and rotates into something that works has had a good day.
The Same Desk, Flipped.
Political positions are stated in directions and aspirations rather than conditions, because a falsifiable claim hands opponents a weapon. The rational adaptation is language precise enough to sound like a commitment and vague enough to survive being wrong.
The closest equivalent to position monitoring is the civil service. The theory is the permanent institutional memory, the risk committee that outlasts the political cycle. The practice is the most sophisticated positional defence architecture in the system. It constructs rationales for positions already taken. It smooths divergence between evidence and policy into the briefing until it disappears.
What makes this more than individual failure is the generational dynamic. A junior official learns not what produces results but what produces approval from a senior official who learned the same lesson from someone above them. The range of options considered narrows with each generation. By the time a genuinely disruptive possibility reaches a ministerial brief it has been smoothed into something manageable, risk-assessed into timidity and presented as the responsible choice. Imagine a trading desk where junior traders are trained exclusively by senior traders whose careers were built on hiding the P&L. Each cohort receives the distilled defensive posture of every cohort before it, presented as wisdom, and reproduces it downward.
The selection pressure running beneath this is what makes it structural rather than incidental. The people who benefit most from the defensive culture are the most senior, which means the institution promotes for timidity rather than drifting toward it. The most sophisticated positional defenders reach the top. The most honest analysts leave or are sidelined. Across decades the institution actively selects against the capacity it most needs and the people sitting at its apex are the proof of that selection, not an aberration from it.
A minister holds a bad position for four years. A civil service culture that defends positions rather than tests them holds it for forty and trains the next generation to hold it for forty more.
The position closes only when external reality makes it untenable. By that point Grangemouth has already closed.
The system’s only stop loss fires every five years at a general election. It is reviewed by an electorate whose judgment is framed by newspaper front pages, leader debate moments and the emotional simplicity that five seconds of attention produces. A minister can make a decision in year one whose damage becomes visible in year six. The stop loss fires after the position has already been transferred to someone else. The clawback never triggers.
The culture runs in precise inversion to fast cutting. Every reversal is a press conference where opponents play the clip. Every honest reassessment is an attack ad. The rational response to accumulating evidence that you are wrong is to generate more reasons why you were right.
The Absent Forcing Function
Investment banking reached workable accountability through failure rather than foresight. Early compensation models rewarded volume, headline profit and annual revenue. They were simple and destructive. Risk moved forward, losses landed later and rewards accrued to those who created the fragility. Structures eventually changed. Multiple metrics replaced single ones, bonuses were deferred, clawbacks were introduced and careers ended when underlying performance deteriorated, even if reported figures held up. None of it came from moral insight. It came from repeated loss.
Government never reaches that forcing function. Losses are socialised, careers survive them and the lesson goes unlearned. The techniques that limit gaming are well known and tested. They get avoided because they attach consequence to authority. The system that needs correcting is the system with the power to resist correction.
The Kitchen Sink
Exogenous shocks are the politician’s ‘kitchen sink’ moment, the corporate move where incoming management loads every accumulated problem (including the kitchen sink) onto a single bad quarter, blames the previous regime and then reports improvement against a floor they set themselves. The shock provides external legitimacy for the write-down. Everything already broken gets attributed to the storm.
Von der Leyen’s Paris speech is an attempted kitchen sink without the precondition that makes it work. The move requires clean hands. She was a minister in the cabinet that made the original decision, then the Commission President who extended it, and she is now attempting to load the write-down onto predecessors she herself was. The accounting does not hold. What it reveals goes beyond personal audacity. A political culture with functioning error correction would produce a credible author of reform, someone who identified the mistake from outside it and paid a price for saying so at the time. This one produced the architect of the error, repositioned at the podium, because those are the only people it grows.
Nobody Is Running Yesterday’s Morning Meeting
Voters say they want politicians who admit mistakes, but revealed preference tells a different story. The politician who says I was wrong is finished by the following morning’s papers. The one who says the situation has evolved and we are adapting, the same thing in defensive language, survives. The media translates honesty into weakness and the public punishes accordingly.
A minister who responds to a formal adverse finding by revising a position needs to be able to present that as competence. The trader walks back to the desk, the position is closed, the loss is in the P&L and the question is what to do today. The P&L is the only record and it already incorporates the cut.
The Translation
The independent error detection platform is the risk committee. A body with a fixed mandate, insulated from the reshuffle cycle, with authority to trigger mandatory policy review when defined evidential thresholds are crossed. Its composition draws from outside the permanent state it is designed to constrain. Its findings are public and attach to named individuals. Personal consequence requires being personal, non-transferable and large enough to change the calculation before the decision - a reduced pension, a permanent public finding of avoidable harm, a bar on public appointments.
The portfolio reckoning is the P&L. Von der Leyen answerable to fifteen years of European industrial electricity costs relative to France. Swinney answerable to the gap between Scottish energy costs and what they could have been. Outcomes tracked honestly against the decisions that produced them.
The pre-committed review trigger is the stop loss. Policy positions lodged at adoption with explicit falsifiable conditions. When those conditions are breached the review is mandatory.
The basket of upstream metrics, energy costs relative to peers, grid connection times, build velocity, real investment per capita, labour productivity in tradable sectors, external balance, operates as the honest aggregated measure that GDP cannot be. Growth supported by imported capability looks healthy until it stops. The external balance component catches it. Floors held independently across components make gaming harder than delivery. Asymmetric downside, where underperformance carries career consequence, clarifies expectations in a way that upside reward alone leaves opaque.
The pre-committed trigger solves something beyond epistemics. The reason fast cutting is politically lethal is that every reversal currently reads as personal failure, the politician who was wrong, caught, forced to retreat. The trigger changes what the reversal is. If the falsifiable conditions are public at the point of adoption, the review firing is a process working as designed. The press conference is not I was wrong about nuclear. It is the conditions we specified have been met and the review I commissioned is reporting back. The embarrassment migrates, as it does on the desk, from the individual to the architecture. A politician who designed a system that caught its own error and corrected it has a record to run on. The voter who punishes honest reassessment is punishing something that no longer looks like honest reassessment. It looks like competence. This is what actually happens when error correction is institutionalised rather than personal. The trader does not apologise for the stop loss being triggered. The stop loss triggering is the point.
No costs to those involved.
This is what the gap costs. Across Europe, industrial electricity now runs at roughly twice the US price - a structural disadvantage compounding across every energy-intensive sector, every investment location decision, every factory that didn’t get built or relocated and didn’t return. The argument for nuclear that Von der Leyen made in Paris was available throughout. The cost of not making it earlier is embedded in two decades of industrial positioning the continent is now trying to reverse from a weakened base. In Scotland, a refinery, a workforce, gone.
In both cases the position closed too late. In both cases the people who held it past the point where the evidence had turned announced the correction, collected the credit and kept the pension.
The loss has never been marked and the cost never recognised in the personal accounts of those involved.
Great article however think I need a month in Dartmoor after reading that
For some reason this brought Jack Grubman to mind. I attended the annual SSB ‘TMT’ conferences in Scottsdale from ‘99-2001. For obvious reasons he stood up in 2001 and completely contradicted everything he had said in the previous 2 years. My nativity at the time meant I was taken aback by his lack of accountability and willingness to admit mistakes, but it taught me these people have no shame. This charlatan ended up in jail. Our politicians walk free.