Confirm, Then Make Irreversible

The standard posture of the advisory industry in a sovereign context is to arrive with a programme. The programme is competent, benchmarked against comparable states, and usually correct on the economics. It is also, in a specific and fatal sense, nobody's.

An imported programme has no domestic author. That is not a complaint about credit; it is a description of what happens when the political cost arrives, as it always does, some months after signature and some months before any benefit is visible. At that moment somebody inside the system has to spend their own standing defending the thing. If the programme belongs to an external adviser, the cheapest available domestic position is that it was imposed, and the cheapest position is the one that gets taken. The reform is then reversed at the first change of minister, and the reversal is inexpensive because nobody present is diminished by it.

The alternative is not a better programme. It is a different relationship to authorship: do not prescribe, confirm — and then build the structure that makes what was already emerging impossible to reverse cheaply.

What confirmation actually requires

Confirmation sounds passive and is not. It is the harder analytic job, because it demands a distinction that most reporting is structured to obscure.

The first task is separating what is genuinely in motion from what has been announced. Those are different populations and they overlap less than the documents suggest. A direction that is genuinely moving has evidence that is boring and structural: budget lines that have already shifted, appointments made to positions that matter rather than positions that are visible, secondary legislation drafted, a department that has begun hiring for a function it does not yet formally hold. An announcement has a launch and a strategy document.

The second task is identifying which coalition is carrying it. Every real direction has one — a set of interests, inside and outside government, whose position improves as it advances. If that coalition cannot be named specifically, the direction is not moving; something is being said about it.

The third task is the one most often skipped: establishing what it would take for that coalition to lose. Which of its members is exposed at the next electoral or budgetary event, which of its gains are reversible, and what a competent opponent would attack first. That analysis determines everything about sequencing, because it tells you which parts of the structure need to be finished before the window narrows.

Irreversibility is a price, not a wall

Then the hardening, and the point most often misunderstood: irreversibility does not mean a reform cannot be undone. It means undoing it has been made expensive, slow and visible, and has been given an owner.

The instruments differ by cost imposed. Statute rather than policy, because a policy is changed by the person who holds the office and a statute requires a legislature. Institutional mandate rather than programme office, because a programme office is closed by a line in a budget and an institution with a statutory function has to be legislated out of existence. Treaty or arbitral commitment rather than undertaking, because an undertaking is renounced and a commitment is breached, and those carry different prices. And capital directed only at what is genuinely movable, because capital deployed against announced rather than actual motion produces a stranded asset that becomes the strongest argument against the direction it was meant to support.

The clearest published illustration of the sequence is the operational independence of the Bank of England. It began as policy: a letter from the Chancellor to the Governor dated 6 May 1997 setting out a new monetary policy framework, with the committee constituted on an interim basis. For roughly a year the arrangement ran on an administrative footing, reversible by the person who had created it.

Then it was hardened. The Bank of England Act 1998 received Royal Assent on 23 April 1998 and came into force on 1 June 1998, and section 11 placed the objectives in statute — to maintain price stability, and subject to that to support the economic policy of Her Majesty's Government, including its objectives for growth and employment.

But look at section 19, which is where the real design lesson sits. The Treasury retained a reserve power to give the Bank directions on monetary policy. Reversal was not abolished. It was conditioned: available only where the Treasury is satisfied that directions are required in the public interest and by extreme economic circumstances, and an order lapses unless approved by resolution of each House of Parliament within twenty-eight days, with an outer limit of three months regardless.

That is what designed irreversibility looks like when it is done well. The escape hatch exists, which is why the structure is durable rather than brittle. What was removed is the possibility of quiet reversal — of the direction being undone by a decision taken inside a building, without a debate, a vote and a named person to answer for it.

The sequencing constraint

The order of those two moves is not interchangeable, and getting it backwards is the common failure. Hardening a direction that is not yet genuinely carried produces the worst available outcome: a statute with no coalition behind it, which is both harder to repeal and easier to ignore. Unenforced law is not a neutral outcome — it teaches every subsequent participant that the instrument does not bind, which raises the cost of the next attempt in the same jurisdiction.

The reverse error is slower and more forgivable. A direction confirmed but never hardened simply reverts, and the cost is the time. That is why the analytic work sits first and the drafting sits second, and why a house that is good at drafting will tend to arrive at the second step regardless of whether the first has been done.

The doctrine underneath

This is disciplined participation in an order the adviser does not command, and the discipline is mostly restraint about authorship.

Countries in transformation rarely lack direction. What they lack is the mechanism to move faster than their own institutional processes allow, and a party with no domestic constituency to satisfy and no public position to protect. That is a narrow function. It is not the design of the destination; it is the removal of friction between a decision already taken and the structure that would make it permanent, and it works because the direction remains attributable to the sovereign throughout.

The test of the work is not whether the structure was elegant. It is whether the reform survived the administration that introduced it, and whether the people who defended it through the expensive period did so because it was theirs.