Somewhere in England this year, a force advertised for a chief constable and received one applicant it considered suitable. Not one outstanding candidate among several — one, full stop. This is not a story about a shortage of ambitious police officers. It is a story about what happens to a leadership pipeline when nobody has been asked to own it for thirty years.

The Police Leadership Commission's report, published Monday and co-authored by Lord Blunkett, puts numbers to what most serving officers already knew. Promotion in England and Wales operates as a postcode lottery: whether you rise depends less on merit than on which force you happen to be in, who your chief inspector happens to like, and whether you can survive a sergeant's exam so badly designed that fewer than half of candidates pass it. The report uses the word "outdated." A civil servant would call it what it is — a filter that screens for exam technique rather than leadership, and has done since long before most of the officers now failing it were born.

Blunkett's language is careful, as it should be from a man who has run a department. He found "outstanding examples of those who have transformed delivery to the public" alongside "extraordinarily worrying evidence requiring profound change." Translate the second phrase properly and it means this: somewhere in the system, for years, people watched leaders fail and did nothing, because nobody's job was to intervene.

That is the actual finding, and it is worth sitting with. Not that individual officers were promoted above their competence — that happens in every institution — but that the mechanism for catching it, correcting it, or even naming it, does not exist. There is no national body responsible for developing chief constables the way NHS England is responsible for developing hospital trust chief executives. There is the College of Policing, chronically underfunded relative to its NHS equivalent; there are forty-three forces, each running its own promotion process; and there are Police and Crime Commissioners, elected since 2012 specifically to provide local accountability, who mostly discovered they had influence over budgets and precious little over how a force cultivates its future leadership.

The honest defence of the current arrangement runs like this. England and Wales already has HMICFRS inspecting forces, the IOPC investigating misconduct, PCCs providing local democratic oversight, and the College of Policing setting standards — four bodies, arguably too many, each with a remit. On this view the problem the Commission describes isn't structural at all; it's a run of bad individual appointments, unfortunate but not systemic, and a fresh national academy risks adding a fifth layer of bureaucracy to a system already thick with them, without fixing the culture that actually produces bad chiefs.

It is a serious argument and it survives contact with about half the report. It does not survive the sergeant's exam, which is a mechanism, not a personality. It does not survive the finding that central funding for leadership development was allowed to wither while the NHS equivalent was protected — a decision nobody announced and everybody who worked in either institution noticed. And it does not survive the testimony, circulating this year among serving officers, of a sergeant with twenty-four years' service walking away because the disconnect between headquarters and the custody suite had become unbridgeable. You do not lose a sergeant of that seniority to individual failings. You lose them to a system that never asked what leadership training he had received since the day he passed the exam he is now watching his own recruits fail.

Here the history is unusually direct. Robert Peel built the Metropolitan Police in 1829 on the principle that the Home Secretary answered to Parliament for how it was led — an accountability line short enough to follow with your finger. It survived, more or less, until the tripartite system of 1964 blurred it between chief constables, watch committees and the Home Office, and it blurred further still when Police and Crime Commissioners arrived in 2012 promising a direct democratic answer and delivering, in practice, a fifth wheel with a mandate nobody much noticed at the ballot box. Each reform added a body. None restored the thing Peel actually built, which was a single point where responsibility landed.

The Commission's answer, a National Academy of Police Leadership, is the correct instinct — consistent training, a credible pipeline, funding restored to something like NHS levels rather than the leftovers it has survived on. Policing Minister Sarah Jones says the recommendations will "shape" the government's reform programme, which is the department's present tense for we have read this and intend to be seen reading it again. Sir Andy Marsh calls it the most comprehensive examination of police leadership in a generation, which may well be true and is also what people say about examinations that produce academies rather than decisions.

An academy is a building, or in practice a website and a curriculum, and it will not by itself answer the only question the report actually raises: who is now responsible for whether the next generation of chief constables is any good, and what happens to them if it isn't? Peel's Home Secretary could be summoned to the Commons and sacked. Name the person who inherits that job now, and you will have done more for policing than the academy ever will.

In 2023, two-thirds of every pound handed to a British political party came from nineteen people. Not nineteen companies, not nineteen unions — nineteen individuals, most of them multi-millionaires, writing cheques large enough that the rest of us writing £20 by direct debit might as well not bother. That number sat in a report from Democracy for Sale for months without moving a single vote in Parliament, which tells you everything about how seriously Westminster treats arithmetic it doesn't like.

Now, suddenly, something is moving. Alex Sobel, a backbench Labour MP, has tabled an amendment to the Representation of the People Bill capping individual donations at £1 million. Simon Opher, another Labour MP, has signed a Liberal Democrat amendment proposing £50,000, falling over time, alongside some form of public funding for parties. And Andy Burnham — the man Labour MPs now discuss the way Kremlinologists once discussed the Politburo, as the person who will actually be running the country by 20 July — has said, in an email seen by the BBC, that there should be a cap. Full stop, no number attached, but a cap.

Here is the part that should stop you mid-sentence. Sir Keir Starmer, the man who will hand Burnham the keys in a fortnight, does not support the cap. The bill that could introduce it is due to finish its Commons stages before Starmer leaves office. Which means the government he currently leads gets to decide, in the next few weeks, whether the country's next prime minister ever gets the chance to implement a policy he has publicly endorsed — by choosing whether to let the bill through, pull it, or let it limp into the Lords in a shape Burnham then has to fight to amend. Nobody voted for that sequencing. Nobody ever will.

There is a serious case against the cap, and it deserves stating properly rather than waved off. Smaller parties and insurgent challengers have historically relied on a handful of committed backers to get off the ground at all — a hard cap can entrench the two big parties' fundraising machines, built over decades, against anyone trying to break in. There's also the underground-money problem: cap the front door hard enough and the cash finds side entrances, dark-money vehicles, unincorporated associations, the kind of arrangement that got Christine Lee's law firm £675,586 into Barry Gardiner's office without anyone quite noticing until MI5 did. A cap that isn't matched by real transparency and enforcement doesn't end influence-buying. It just makes it worse at hiding.

That case is real, and it still doesn't survive contact with the number this piece opened on. Nineteen people. Two-thirds of the money. A £1 million cap, which Sobel is proposing as the serious reform, is not a modest concession to that problem — it is a cap set at a level most people will never see in a lifetime of payslips, on donations from individuals whose current giving already runs into the tens of millions across an election cycle. Opher's £50,000 is closer to something a reader can measure against their own life: it is roughly what a couple on median full-time earnings brings home before tax across two years, which means even at the tighter figure, a single donor can still outweigh several hundred ordinary voters combined, legally, every single cycle.

And notice what both amendments leave standing. Sobel's cap applies only to individual donors — collective organisations, meaning trade unions, Labour's largest funding bloc for a century, sit entirely outside it. So the proposed reform caps the Conservative Party's route to money and leaves Labour's largely untouched, which is either a coincidence or isn't, and I'd invite you to guess which. This is the trick that runs through nearly every donations-reform conversation Westminster has had since Hayden Phillips first proposed a £50,000 cap back in 2006 and got nowhere: everyone agrees the other side's money is the corrupting kind.

The housing secretary, Steve Reed, has already announced a moratorium on crypto donations and a cap on money from Britons living abroad — genuinely useful, narrowly targeted, and entirely compatible with leaving the nineteen-donor problem untouched. It closes the loophole nobody's rich uncle uses and leaves the one the rich uncle does use exactly as it was.

What's actually being decided here is not whether Britain gets a donations cap. It's whether the decision gets made by the government currently in office, on its way out, with no appetite for it — or by the one about to arrive, who says he wants it, on a timetable set by whichever whip decides how fast a bill moves through committee. That's not a question with an answer in any manifesto, because no manifesto describes the mechanism. It was decided in the choreography of a handover, by officials scheduling Commons business around a leadership transition nobody outside the parliamentary Labour Party had a ballot on.

Burnham may get his cap. It may even bite. But the fact that whether nineteen people keep buying two-thirds of British democracy depends on which side of 20 July a bill happens to land is not a triumph of reform. It's a demonstration of exactly the problem the reform is meant to fix.

At four in the morning on 19 September 2014, Mary Pitcaithly stood in front of the cameras as Chief Counting Officer and delivered the line the whole apparatus of the referendum had been built to earn: "I am satisfied that all counts were conducted properly." Hundreds of independent observers had watched every count. Officers from Police Scotland had stood in every hall. International monitors had flown in to certify that all 3,623,344 ballot papers had been handled honestly. Turnout was 84.6 per cent, the highest in any Scottish election since women got the vote in 1918. It was, by any reasonable measure, one of the best-scrutinised democratic exercises this country has run.

The votes were audited to destruction. The money was not audited at all.

Yes Scotland Ltd, the company Alex Salmond set up in May 2012 to run the official pro-independence campaign, disclosed in April 2013 that it had taken in £1.6m in donations, more than a fifth of it from a single euromillions-winning couple, plus £342,797 from the SNP itself for start-up costs. It fought the campaign, lost it, and then did what campaign companies do: it went dormant. No activity since 2014. A run of quiet accounts, the most recent published last year, showing a balance of zero.

That is where the story should have ended. This week, twelve years on, a Police Scotland complaint says it did not. The complaint comes from David Henry, a former SNP branch secretary whose earlier concerns about party finances triggered Operation Branchform, the investigation that eventually put former SNP chief executive Peter Murrell in prison for more than five years for embezzling £400,000. Henry's new dossier concerns a note in Yes Scotland's 2014 accounts: income of £1,524,998 that, he says, does not reconcile with what follows. "Major corrections appeared in the 2015 accounts," he told the Sunday Mail. "Nothing adds up." Police Scotland confirms only that "inquiries are ongoing," which is the institutional equivalent of a shrug with a case number attached.

The SNP's defence, offered promptly, is that Yes Scotland Ltd was "an entirely separate organisation" from the party. True in the sense that a subsidiary is separate from its parent, and less true in the sense that the party founded it, staffed its launch, and had its books signed off by SNP officers. The defence deserves a fair hearing, because the separation argument is not absurd on its face. Campaign companies exist precisely so that a single-issue, cross-party effort can raise and spend money without dragging every donor into party membership. Yes Scotland always insisted it was broader than the SNP, drawing on Business for Scotland, Women for Independence, a scattering of actors and a former Bond. Better Together was structured the same way. There is nothing sinister, in principle, about a referendum vehicle that dissolves once the referendum is over.

But dissolution is exactly the problem. A company can go dormant. A question cannot. Yes Scotland folded its books, filed its zeros, and took the answer to "where did £1.5m go" into administrative hibernation along with it — beyond the reach of journalists, of the SNP's own later scrutiny, and, until Henry walked into Gayfield Square police station with a dossier, of the police. Nobody audited the audit. The company that owed an explanation had, by the time anyone thought to ask for one, no obligation left to give.

Trace the accountability chain and it runs out where these chains always run out: at a company registrar's filing cabinet. Who signed the 2013 accounts. Who authorised the corrections in 2015. Who decided that a company holding £1.5m of public donations for a national referendum could report a zero balance a year later and call that a filing rather than a confession. These are not unanswerable questions. They are simply questions nobody with the power to ask them chose to ask, for a decade, until a man with a grudge and a spreadsheet did it for them.

There is a reason this keeps happening in the same shape. Murrell embezzled from a party with members, conference votes, and at least the theoretical possibility of internal challenge. Yes Scotland was a company with directors and no electorate at all, which raised more money in eighteen months than most parliamentary candidates see in a career and then dissolved before anyone had reason to ask where it went. Britain's electoral law regulates parties reasonably well and regulates campaign companies barely at all. The gap between the two is where the £1.5m sits, presumably, still.

This is not a Scottish curiosity. The current Westminster argument over donation caps and undisclosed benefits — over whether a wealthy individual can fund a party through a shell, a loan, or a gift in kind that never quite counts as a donation — is the same argument at a different postcode. The Electoral Commission can fine a party. It has essentially no purchase on a company that campaigned for a single vote and then obligingly stopped existing. We built one of the best-observed ballots in our history and left the wallet on a park bench.

The count was watched by thousands. The money was watched by nobody, and the difference between those two facts is the whole of this column in miniature: we have become extremely good at auditing the parts of democracy that are cheap to audit, and have stopped bothering with the parts that cost something to check. Scotland asked, on 18 September 2014, whether it should govern itself. It got an answer. The country is still waiting for the answer to the smaller, ruder question — where the £1.5m went, and who was ever supposed to make sure we'd know.

Two years ago this week, Labour won 412 of the 650 seats in the House of Commons, the largest majority any party has assembled since Tony Blair's in 1997, achieved against a Conservative Party in a state of genuine institutional collapse. Two years on, almost to the day, Keir Starmer has resigned as prime minister, Labour is choosing his successor, and Morgan McSweeney, the strategist who built the machine that won the landslide and then followed Starmer into Downing Street as chief of staff, has told whoever will listen that the government was not ready to govern.

There is a case for him, and it is worth stating plainly before it is dismissed. Labour's rise was sharper than anyone in the party had planned for. A fragmented state, inherited mid-collapse, does not come with an instruction manual. The "missions" framework Starmer launched in February 2023 was deliberately built around outcomes rather than delivery mechanics — clean power by 2030, growth to the top of the G7 — on the theory that fixating on means before you hold office produces manifestos full of detail and governments full of excuses. Slow, careful, mission-led government is a defensible strategy. Nobody sensible expects a change of government to run like a factory changeover.

But the manifesto did not promise patience. It promised foundations. "The starting point for delivering these missions," Labour's own governing document said, "is to ensure the foundations of good government are right." Translated out of manifesto-speak, that sentence means: before we do anything else, the machinery will work. Two years later, the man who wrote the campaign that sold the sentence is confirming, in public, that the machinery did not work, and treating the confession as insight.

This is not an excuse, and it should not be received as one. Starmer became Labour leader in April 2020. He had four years, not four weeks, to build a transition operation, and he was not improvising in the dark. The Institute for Government had been running a podcast series called Preparing for Power since February 2024, publicly interviewing former officials and advisers about the machinery of taking office, because the problem of transition readiness is old, well-documented, and entirely foreseeable to anyone paid to foresee it. A party that spent four years drafting five missions apparently could not spare a fortnight for a delivery plan, and nobody has yet lost a job over the omission.

Which returns us to the only question this column ever asks: who decided this? Not who decided the missions — that was Starmer, in public, campaigning on them. Who decided that readiness itself was somebody else's job? McSweeney ran the campaign and became chief of staff; if anyone in the modern Labour Party held formal responsibility for turning a landslide into a functioning government, the organisational chart points at him. His admission is therefore an unusual kind of testimony: the man in charge of preparation confirming the failure, then handing the postmortem to a leadership contest he did not resign to enter.

Compare 1945. Attlee's government took office into a country that was bankrupt, rationed, and owed America more money than it could see a way of repaying, and it built the National Health Service and 200,000 houses a year regardless. It could do so because the civil service had been planning reconstruction since Beveridge reported in 1942, three years before the election that let anyone act on it.

The preparation was not glamorous. It was committees, drafts, arguments about hospital boundaries, done in wartime on no budget, years ahead of the vote that mattered. Labour in 2024 had the same lead time and a functioning peacetime civil service to plan with, and produced, according to its own chief of staff, insufficient foundations.

None of this would matter much if someone had been held to it. A failed transition is not a scandal in itself; every government since Attlee's has underdelivered against its own promises. What makes this one an accountability story rather than an ordinary tale of governing difficulty is the total absence of a name attached to the failure. Nobody has been sacked. Nobody has resigned over it, and Starmer's own departure has been framed around other pressures entirely, which is itself instructive. The admission arrives as diagnosis rather than confession, delivered by the one man best placed to have prevented the disease and least willing to say so plainly.

That is not repentance. That is positioning — an attempt to relabel a personal and organisational failure as a systemic inevitability, arriving at the precise moment a leadership contest is choosing who inherits the blame. It is a neat trick, and it will likely work, because the members voting for Starmer's successor have no mechanism to ask McSweeney to account for the last two years and every incentive to look forward instead.

Two years after 412 seats, the government's own former architect says the plan was never built, and the man saying it is still in the building.