David Davis had a suspicion. The Brexit secretary realised he was being increasingly isolated and sidelined by the machine in the first quarter of 2018. “Something happened. I don’t know what. Something happened.”
He was right.
In December 2017, the UK had agreed the so-called Joint Report with the EU – the document that confirmed the initial shape of the backstop. It had been a difficult, gritty negotiation, and had damaged the relationship – such as it was – with Ireland. Olly Robbins, the prime minister’s sherpa, had a good relationship with his Irish opposite number, John Callinan.
But Davis had a particularly poor relationship with Simon Coveney, the Irish foreign minister and Tánaiste. According to a friend of the Brexit secretary, Davis “loathed” him. (Davis himself told me, rather diplomatically, that he thought their meetings “went OK”. )
The Irish had a problem with him, too. Just after the Joint Report was agreed in December 2017, Davis went on the Andrew Marr Show, a Sunday morning TV politics show on the BBC. Yes, the backstop was bad, he said, but it was only a “statement of intent”. Not a legally enforceable document. This brief interview is still brought up spontaneously by Irish and Commission officials.
There was also a particular problem in getting the prime minister and Leo Varadkar, the Taoiseach, to converse. Their relationship is said to be “spiky”, “difficult” and “cold”. Both are personally a little socially awkward.
Three people with knowledge of their discussions in London and Dublin spontaneously used the expression “on the spectrum” when describing their interactions – a reference to autism. (A friend of the prime minister puts it differently: Varadkar is a “cooler character” than his predecessor, Enda Kenny. He and May are of “different generations, different outlook on life and politics”.)
In order to repair the Irish relationship, the UK set up a back-channel to Dublin. The Irish government was asked to stop talking to Davis and to deal instead with David Lidington, a former Europe minister.
As May’s de facto deputy, Lidington was also Coveney’s opposite number as Tánaiste. Lidington and Coveney regularly met and spoke by phone – “weekly”, according to one account. On their first meeting, they had a 20-minute session alone. “Oh, yeah, they text all the time,” one Irish diplomat told me.
But Davis was never informed. “No, I didn’t know that,” he told me. There was sheepishness in both Dublin and London when I asked about it; while the Lidington-Coveney talks became increasingly open, they had been agreed without telling Davis. No one ever felt the need to tell him about them, or explain why they were happening.
This was just a small microcosm of a trend. Increasingly, DExEU and the Brexit ministers were sidelined. One official likened it to The Truman Show, a film about a world built for a TV show, where only the protagonist does not know his reality isn’t real.
There was, another official said, “DExEU I” and “DExEU II”; officials maintained the pretence that the department worked for its secretary of state, but it really worked for Olly Robbins. Ministers were seeing his shadow everywhere.
Between December 2017 and mid-2018, a senior official told me that Davis was “promised over and over and over again, it would be different this time. Olly would be told to include him, to talk him through the PM’s thinking”.
It never was different. And the deterioration of the relationship between Davis and the prime minister gained its own momentum. Leaking from both sides became a problem; each side names officials or special advisers they blame for disclosures. Despite an apparently genuine desire by the two central figures – May and Davis – to make it work, they were drifting apart.
A close ally of the prime minister said: “Their approach to negotiations is utterly different. The PM is somebody who just believes you never leave the table, you just grind on. And that’s how you get things done. And you wear people down through sheer resilience, whereas DD would much rather say: ‘…bang something down on the table and you walk away and wait for them to come to you’.”
It was doomed.
The extent to which DExEU’s ministers were in the cold was visible in the department.
In March 2018, Donald Tusk, president of the European Council, gave a speech. Britain was officially leaving the customs union. It was officially leaving the single market. So what did that mean?
“The only remaining possible model [for Brexit] is a free trade agreement. I hope that it will be ambitious and advanced – and we will do our best, as we did with other partners, such as Canada recently – but anyway it will only be a trade agreement.”
That is, a distant relationship. In truth, it suited a lot of leavers, including the Brexit ministers. In fact, it was the unanimous choice of all the DExEU ministers. Suella Braverman joined the department as a minister in January 2018.
She said of a Canada-style free trade agreement: “Ministers did talk about it at length in DExEU and we had a unanimous view that that would be the preference, but it didn’t seem to land or be taken by Number 10 … The line was, ‘we’re not taking an off-the-shelf template’ … I was never allowed to say ‘Canada’s a good solution’ publicly.”
Steve Baker, who was in the department for around a year from mid-2017 told me: “I started wanting to talking about that Tusk offer. It was taken out by officials working for Olly Robbins. My staff would say: ‘Oh, Number 10 have taken it out’. I’d never accept that. ‘Who at Number 10?’ And it was Catherine Webb, working for Olly Robbins. God bless her, I’m sure she’s very talented and a nice person. But what right did Catherine Webb have to take this out of my speeches? I’m outraged.”
I asked someone who worked with Webb and was told: “She was doing God’s work. That Tusk offer did not apply to Northern Ireland. It would have meant a border in the Irish Sea. It would have been back to square one. She was absolutely right to stop that. That was not government policy.”
By 2018 the frustration of the Brexiters was becoming very clear. In March of that year the UK agreed to a transition period (formally known as an “implementation period”). This would mean that if the UK were to agree a withdrawal agreement, it would get another 20 months or so of being treated like a member state so that it could prepare for departure, and negotiate a detailed settlement for the relationship to follow.
To some Brexiters, this was common sense. The prime minister had introduced the plan in late 2017 in her Florence speech. The EU had consented to begin discussion on it once “sufficient progress” was made. To other Brexiters, though, it seemed like a concession. The UK’s initial ambition, gradually chipped away during the process, had been to conclude a trade agreement with the EU alongside the withdrawal agreement before the two-year clock was up.
The original Article 50 notification from May had stated: “We recognise that it will be a challenge to reach such a comprehensive agreement within the two-year period set out for withdrawal discussions in the treaty. But we believe it is necessary to agree the terms of our future partnership, alongside those of our withdrawal from the EU.”
Many negotiators expected there would have to be a further period of transition. At the start, officials say, they were continually pushing back on this two-year timetable as unrealistic, because trade talks typically take much longer. But such gloominess was contentious; Sir Ivan Rogers resigned as Britain’s lead official in Brussels in January 2017, following the leak of a memo he had written which had made the argument that the trade negotiations could take years – and still fail.
Suella Braverman, a Brexit minister in 2018, said that David Davis adopted this position “because he’s loyal … and to reassure remainers and business that the implementation period is about certainty and predictability”. It was, she said, one of the first compromises the eurosceptics made. For them the transition was something that could be tolerated if it got them to a deal they could be happy with. That, though, looked increasingly unlikely.
At the start of 2018, after the Joint Report, one question remained unanswered: what did Britain actually want from the EU? Having raced to get a “sufficient progress” decision from the EU, there was then seemingly little urgency in answering this question.
Permanent secretaries all over Whitehall complained to Sir Jeremy Heywood, the cabinet secretary, about this problem. There was concern that either Robbins was not making sure decisions were made – or he was being so secretive about the decisions that no one knew.
Heywood found himself in an impossible position, trying to manage this cat’s cradle of conflict inside government while, at the same time, dying. Immediately after the general election in 2017, he was diagnosed with lung cancer. He would, eventually, take a leave of government in June 2018 – but even before that he was slowly withdrawing.
Particularly following the departure of Fiona Hill and Nick Timothy, May’s two chiefs of staff, in mid-2017, the gradual departure of Heywood was important. He set a tone. He was both an architect of mechanisms that were intended to help the prime minister avoid making choices and a critic of her indecision.
The impact was “huge”. As one close colleague put it, “Britain’s finest and most creative policy and political problem solver was steadily removed from the scene”. The prime minister went to see him in his south London home to seek his counsel, but his creativity and ability to reframe problems was day by day being drained from government at a time when Britain needed to make big decisions.
He was disappearing from view just as key questions remained to be resolved – on regulation and customs.
On regulation, the UK had been clear it would not be inside the single market. Being inside it would mean, for example, that British banks could operate wholly freely across the whole EU from London, and all goods would be deemed to meet EU standards.
But it would mean being inside the European Court of Justice and having no controls on EU migration. In any case, there was strong demand from Brexiters to permit deregulation: they did not want to retain the EU’s rulebook.
But this would cause problems for UK companies: products might need to be re-tested in the EU, or companies might be required, in effect, to set up EU bases. There might be EU-UK border hassle from regulatory checks.
Potential solutions ranged widely in difficulty and effect. Would it be possible, for example, to agree structures that would allow British toys to be sold in the EU, without needing extra product testing inside the EU?
“There will be areas where we want to achieve the same goals in the same ways, because it makes sense for our economies,” May said in her Florence speech in September 2017. But she also proposed that Britain might want to diverge from EU rules but still have UK goods deemed acceptable.
The City in particular hoped to benefit from this sort of approach. In March 2018, the prime minister gave a speech in the City of London saying that financial services rules needed mutual recognition: “Our goal should be to establish the ability to access each others’ markets, based on the UK and EU maintaining the same regulatory outcomes over time, with a mechanism for determining proportionate consequences where they are not maintained.”
There was already a process for this – “equivalence” – but it is a one-sided process. It does not cover the full range of financial services, can be withdrawn at any point by the Commission and would make Britain a rule-taker. The EU could change direction, and Britain would need to follow it. Britain wanted something a little more even-handed.
On customs, Britain had to square several circles. The EU position was that any deal had to maintain a fluid border between Northern Ireland and Ireland, or the backstop would be triggered and the six counties of Northern Ireland split off from the rest of the UK.
The Commission also had a dim view of British customs more generally: the EU had started to raise concerns about the UK’s enforcement of value-added tax rules in relation to Chinese textiles. (This was universally regarded as a ruse in London, intended to weaken the UK position.)
Domestically, Britain’s economics ministries wanted a strong link to Europe. The government’s own analysis was that being in a customs union, for example, with the EU would definitely be better for the UK economy than being outside.
Yes, it would mean no significant trade deals. But their economic impact, in truth, would always be extremely small, and a customs union would be valuable. It cuts border hassle and makes sure all UK goods automatically attract no tariffs.
A lot of Brexiters thought that an independent trade policy was worth fighting for – on sovereignty grounds, even if not on economic ones. That was the context in which political attention returned to two options on the docket from a paper published by the UK government in the summer of 2017.
One idea was known as “max fac” (for “maximum facilitation”) and the other as the New Customs Partnership (or “NCP”). These were policies designed to allow a close enough relationship with the EU that there could be an open border in Northern Ireland – while, at the same time, permitting a stand-alone trade policy.
Max fac was the idea that you could have divergence from the EU on customs – so Britain could cut tariffs as part of an arrangement with, say, China. That would require some sort of process by which goods entering Ireland from Northern Ireland could be checked for Chinese imports – in short, a border on the island of Ireland.
Max fac advocates argued that you could, however, make an invisible border. Giving waivers to small traders would take care of 80 per cent of traffic, with trusted trader schemes for big traders doing much of the rest. High-tech solutions could resolve any remaining checks.
Braverman argues that there are plenty of tried and tested examples around the world of customs and checks being carried out away from the border, including the EU’s own Economic Operator Regime – a trusted trader scheme for companies doing lots of business over the border.
Some of the technology proposed was a little unusual. Private companies would drift in with schemes of varying plausibility. Officials would also theorise about how things might be done. Chips on goods, trackers on cars, chemical markers in milk. One official, referring to a private sector idea, asked me: “Has someone told you about facial recognition for pigs?”
There were, however, three broad problems with max fac.
First, both Ireland and EU had rejected this already. Ireland said no to technological solutions to border issues in January 2017. And max fac plans relied on exemptions for small traders, which the EU had also rejected in 2017. They would not allow a leaky border.
Second, the Joint Report of December 2017 had stated that there should be “avoidance of a hard border, including any physical infrastructure” – a serious problem for most max fac schemes. But the report also said there should be no “related checks and controls”.
Max fac plans almost all assumed, in the words of one official working on them, that “you would need some checks away from the border at that time for large businesses”.
According to sources in London and Brussels, this language about checks, so unsettling for some Brexiters, was introduced in papers sent to Ireland and the European Commission by Britain in order to stop proposals that, in effect, just moved the physical border.
The third problem was while you can very easily build a system that makes it easy for people who want to comply, it is not clear how you can have a system that will stop a van driver who does not want to comply, unless you run checks. Borders are hard to do away with. They are the only place where officials know that goods, a driver and paperwork will all be in the same place at the same time.
This side of the argument also had the support of a cottage industry of pro-Brexit think-tank reports, produced largely by the same few people at the London-based Legatum Institute and Institute of Economic Affairs. “They were important”, one of the UK’s negotiators told me, “in maintaining the plausibility of the superiority of that fiction over the other fiction.”
The New Customs Partnership had its own awful problems. The idea was that Britain would administer the EU’s tariffs at its borders as though it was in the EU customs union. At the start, indeed, it would just be an ordinary customs union.
But, later, Britain would be allowed to arrange a trade deal with, say, China, that would permit lower tariffs on Chinese imports so long as you could prove the goods stayed in the UK. The ambition was quite a complex net of processes designed to allow different tariff regimes in the UK and EU, but no need for customs checks between them.
A lot of work was done on the NCP, the proposal that once had been known as “hybrid”. It evolved – but it started off with technical problems about the tracking of goods and legal issues around the collection of tariffs.
It also represented a competitive problem for the EU. The Commission fundamentally believed that if the UK wanted customs union-level access to its market, Britons should not be allowed to undercut the EU with trade policy that made it cheaper to import from other nations. This could lead to British companies getting an advantage by being able to import more cheaply. That is the point of customs unions. In the Commission’s view, the NCP raised “level playing field” issues.
There was serious scepticism among senior civil servants about these plans. I asked a range of people directly involved at what point HM Revenue and Customs officials stopped taking these two ideas seriously. Four out of five said 2017.
Most officials felt they were, to use the words of one Revenue official, “keeping a corpse warm” on both the NCP and max fac. Despite this, the work continued.
In April 2018, two senior officials, Jim Harra and Simon Case, gave a presentation to the European Commission, explaining the two options. The feedback was grim. “They got laughed at,” one Commission official told me. “It was, by all accounts, appalling.” At the time, the Telegraph’s Peter Foster recorded a “forensic annihilation”.
The Commission’s view was that max fac was absurd: a leaky border for goods coming into the EU was out. They were willing to discuss facilitation more broadly – how to make things smoother – but max fac could never be a solution.
On the New Customs Partnership the Commission was also brutal. Their view was that it was also a non-starter. This verdict was faithfully transmitted back to London.
Our negotiator said: “NCPers heard max fac dead. Max fac-ers heard that NCP dead. By this stage of events, everyone [was] looking for evidence they needed to make their version of the answer win.”
NCP, though, was left “marginally less dead … Both options don’t work, but one can be turned into something”. Two officials, one in Brussels and one in London, told me that the EU view was that the NCP could be a ramp into a customs union.
Another UK official said the Commission left the impression that there could be a conversation which started with the NCP because, unlike max fac, it acknowledged that the single market needed leak-proof borders. But, again, the fixes required to make the NCP work in practice all led it closer to being a de facto customs union.
Ministers and officials have told me that the version presented to the Brexit cabinet committee undersold how much the Commission disliked both options. The message to them was both will be very difficult, but they say they are willing to talk about the NCP. And the likely future path of talks on NCP – leading closer to a customs union – was not made clear.
An official said: “The reality is [the Commission was] telling us they hate both of them [but] people were presenting them in a certain way … That’s been one of the most concerning parts of the process.”
It was an important moment for this message to get through. Cabinet ministers were locked in arguments about the two proposals throughout the first half of the year. This would only be brought to a close when the government announced that the cabinet would come to a decision about the final relationship at a meeting on Friday 6 July.
The venue was to be Chequers, the prime minister’s country retreat. The atmosphere beforehand was rather tense.
As one cabinet minister explained to me: “No meaningful discussions were taking place with the EU. If [we] didn’t reach an agreement by November 2018, the assumption was we would leave without a deal.”
On the Monday before Chequers, on 2 July, BBC Radio 4’s Today programme started with some news: Nick Robinson, the host, cheerily announced: “Downing Street say they have devised a new post-Brexit customs plan to try to end divisions among senior ministers.”
Chris Mason, a BBC correspondent, reported: “Downing Street hopes it’s found a way out of the bind on customs – and a significant part of the solution to keeping an open border with the Republic of Ireland.”
According to friends of Davis, this was the first that the Brexiters knew of plans that were being drawn up elsewhere in government. The Brexit secretary rang the prime minister that day, but was told: no, there’s no new plan.
The next day, however, Harra, the HMRC official who had taken customs plans to Brussels, briefed Davis on a new proposal – the set of ideas that would become known as “Chequers”.
The scheme had two elements. The first was a hybrid customs plan – something similar to NCP. But it came with tweaks that meant that in reality it was much closer to a customs union (Sir Ivan Rogers has called it “Schrödinger’s customs union”).
The second was a “common rulebook” by which the UK would sign up to a large number of regulations on goods standards, particularly in highly standardised sectors such as chemicals and medicines. The ambition was to remove regulatory sources of border friction.
On the Thursday, when other cabinet ministers were given the documents, Boris Johnson, the foreign secretary, rang through to Davis. According to one witness, his opening verdict was: “David, have you seen this fucking bollocks?”
The Brexiters were losing.
At Chequers, there was an argument – and there was a victory. After some half-hearted discussion, the cabinet backed the plan.
Davis says the whole thing was “rigged”. During an informal discussion before the cabinet’s ruminations, “Boris got quite emotional, but I stayed out of it”. Johnson’s concern had been, above all, the common rulebook.
What of financial services, so central to the UK economy? A paper that emerged at the end of the Chequers process stated the UK accepted that “future determinations of equivalence would be an autonomous matter for each party”. In practice this meant it would be up to the EU to define equivalence.
This marked an important moment for the Treasury leadership, which had retained dim hopes of very close alignment with the EU for services. But, as the political impossibility of that became clear, the department began to perform what one official called a “pivot”. In effect, Philip Hammond, the chancellor, gave up on secure access for finance in the hope of a better deal for goods.
The Chequers summit was tough on attendees. As one official put it: “Ministers had to hand in their phones, had no officials, and [most] got to see the document only on the day. This was a hugely detailed document that de-prioritised whole sectors of the economy hidden within language that looked like minutiae and was never understood by [cabinet] ministers even on their portfolios.”
But a deal was done.
On 8 July, Davis had finally had enough. He resigned.
The next day Johnson followed him out.
There is already a reading of history which sees the civil service as having led the prime minister to this place.
Reflecting on her behaviour in this period, a minister told me: “It’s Olly. It’s definitely Olly. And the prime minister is a tacit appeaser. She doesn’t really have any views of her own … that aren’t impressed upon her by Olly and Nick [Timothy] and civil servants, right. She’s the puppet.”
In one version of this reading of the past, the decision to concede on sequencing of the talks and on the backstop were intended to make a clean break impossible. Then, this argument goes, the mandarins colluded to press the cabinet onto rails that led to a de facto customs union.
The truth seems to be simpler.
There was no master plan; the civil service went along with what the prime minister asked for. First, on Northern Ireland. Then on customs. She wanted something neither in the customs union nor out of it. Officials bent to the will of the prime minister in persisting with the hybrid idea. Furthermore, I have not found any proof that Robbins ever acted without her approval.
Heywood privately believed that there were “enormous” amounts of evidence that the two proposals considered by the government in early 2018 were non-starters. But ministers are meant to decide, not officials, and he observed the proprieties. So he defended the hybrid, from start to finish, against other civil servants who wanted it killed it in favour of a more crystallised decision.
Stress inside the machine rose. By early 2018, a close ally of the prime minister said: “There was this sense that … we have to come to a decision. Olly was getting increasingly fed up that you weren’t getting … decisions. There had been a growing sense of frustration.”
Around this time, in March and April of 2018, Heywood told people on two occasions that he wanted to push the prime minister into accepting a customs union. But he failed.
Chequers might well, in the pressure of future negotiations, just collapse into a customs union. It is designed to start as one. But the prime minister has persisted with the hybrid – even though it makes negotiations more complex. If she would accept a clean customs union, the backstop would be a smaller issue, and the future partnership talks much clearer. The desire to preserve Chequers as something separate from a customs union is the cause of a great deal of pain.
There is talk in her inner circle about having to persuade her to move. A close ally also noted she has long tended to acquiesce to senior officials. But she is dogged. She is her own person. Someone who has worked closely with the prime minister recently said: “She’s been on a journey … She’s almost on Earth.”
All Photographs Getty Images