Leak inquiries are so rarely successful – I can only stand back and admire Mark Spencer’s audacity

David Cameron used to tell a good story against himself. The cabinet secretary, the late Sir Jeremy Heywood, was a little concerned about a leak of sensitive information about national security. “A little concerned” is mandarin-speak for “an absolute disaster that could lead to World War Three”, so Sir Jeremy recommended to the prime minister that he authorise a full-dress leak inquiry, as there were only a small number of people who had access to the information. 

As Sir Jeremy explained the story a light of recognition crossed Cameron’s face. “Ah,” he said. “That might have been me. You see, I was sitting next to someone at dinner and…”

Leak inquiries are rarely so successful. It would seem that Boris Johnson might have suspected Dominic Cummings and Lee Cain, his departing advisers, of leaking news of his decision at the end of October to order a second lockdown, but the inquiry into the source of the story, which included advisers having to hand over their phones, has not yet come to a conclusion. 

Even the sacking of Gavin Williamson as defence secretary by Theresa May last year – for, in her view, leaking a decision by the national security council to allow Huawei to help build the 5G network – wasn’t based on definitive evidence, and he continues to protest his innocence. 

Now, however, we have a Grade I listed example of how to track down the source of a leak. Civil service legend is full of stories of top secret documents being retyped with tiny changes designed to identify which copy has been leaked, but I have never heard of someone actually doing it – and it working. 

So hats off to Mark Spencer, the Conservative chief whip, who pulled off a magnificent sting this week. He wrote on Monday to the government’s parliamentary private secretaries and the party’s vice chairs, warning them against leaking. 

PPSs are MPs appointed as unpaid hangers-on to ministers. The roles are the first rung of the ladder of promotion, but they are also useful to the chief whip, because PPSs have to vote with the government and abide by collective responsibility. When Theresa May ran out of PPS posts to appoint, she created a number of vice chairs of the Tory party, which allowed even more MPs to be kept in line. 

It is almost part of the unwritten British constitution that warnings against leaking are leaked. At his first cabinet meeting as prime minister, Tony Blair urged his colleagues to maintain discipline and not to leak. His comments were immediately reported by the media. What was surprising about Spencer’s letter is that it took three whole days to appear on the Guido Fawkes website. Immense self-restraint all round. 

But each of Spencer’s letters contained slightly different wording, and so the whips’ office was able to identify which one had found its way into the public domain. Andrew Lewer, MP for Northampton South and a PPS to Kit Malthouse, the home office minister, was duly relieved of his post (while having denied leaking the document) in a denouement worthy of Poirot. 

Indeed, it was better than Poirot because it was real life; it was worthy of Coleen Rooney’s operation to discover who was giving stories from her Instagram account to the press. That case is still in the courts, so we have to be careful. Suffice to say that the version according to Rooney is an absolute humdinger, but is vehemently denied by the accused, and now the Tory chief whip has provided a sequel. 

Lewer has suggested the leak could have come from his staff, which is less gallant but more plausible than the usual “I was hacked/burgled/stitched up”; he protested to Emilio Casalicchio of Playbook: “In nearly 20 years of elected office I have never leaked to the press.”

Of course, journalists ought to be in favour of leaking, but we are also in favour of good stories, and Spencer’s sting was so well crafted that we should only stand back to admire the quality of the work. 

And we should remember, for any important story in the public interest, not to reproduce the whole document so that it can be identified.