Almost exactly a year ago I wrote a slightly tentative blog about small boats crossing the Channel. Something odd appeared to be happening. In August 2025, the number of people arriving in Britain by small boat had fallen by around 40 per cent compared with the previous month. It was the lowest August figure for five years.
I was cautious. “One month’s data doesn’t make a trend,” I wrote. “But it’s worth watching.”
So I watched. A year later we can probably remove some of the caution. In the twelve months from September 2025 to August 2026, 26,837 people arrived in Britain by small boat. In the equivalent twelve months a year earlier, the number was 44,767. That is a fall of 40 per cent, and around 41 per cent below the 2022 peak.
Something really has happened. The interesting question is what.
A problem Britain helped to create
There is a tendency to talk about small boats as though one day thousands of people in northern France simply decided to buy dinghies and sail to Kent. But the explosion in Channel crossings coincided with a very significant change in Britain’s relationship with the European asylum system.
Before Brexit, Britain participated in the Dublin Regulation. The principle was fairly simple: an asylum application was supposed to be considered by one participating European country, rather than allowing someone to make repeated applications in different countries. More importantly, Britain participated in Eurodac. When someone claimed asylum, their fingerprints could be checked against a European database. If they had previously made an asylum claim elsewhere in Europe, Britain could know about it.
This wasn’t theoretical. In 2020 Britain made 8,502 requests under the Dublin system to transfer asylum seekers to other European countries. The Home Office said the sharp increase was partly caused by small-boat arrivals and that many of those arriving had previously applied for asylum in an EU member state. The number actually transferred was much smaller, particularly during Covid, but that misses an important point.
The system provided information.
Someone who had unsuccessfully claimed asylum in France, Germany or Italy couldn’t necessarily arrive in Britain, throw away their paperwork and simply begin again as an entirely unknown person. Their fingerprints hadn’t gone overboard in the Channel.
Then Brexit happened. On 31 December 2020 Britain left the Dublin system and lost access to Eurodac. We never replaced it with anything equivalent. The current Home Office guidance is admirably straightforward about the consequence: Britain no longer participates in Eurodac, so evidence about an asylum seeker’s previous time in EU countries has to be established by other methods.
We still fingerprint asylum seekers. What changed was what we could compare those fingerprints with. This is worth remembering the next time someone complains that migrants deliberately arrive without documents. Some undoubtedly do, and documents can be destroyed or concealed. That happened before Brexit too. The difference is that before Brexit a missing passport did not necessarily prevent the British state discovering that someone had already encountered the asylum system in another European country.
We chose to give up that capability.
Then we got tough
Having weakened one of the administrative mechanisms for managing asylum, successive Conservative governments responded to the resulting problem in a very British way. They announced that they were getting tough.
Priti Patel produced the New Plan for Immigration in 2021. The Nationality and Borders Act followed in 2022. Then came Rwanda. Then the Illegal Migration Bill in March 2023, while provisions from the previous Nationality and Borders Act were still being brought into force. That became the Illegal Migration Act in July. The Supreme Court ruled the Rwanda scheme unlawful in November, so there was a new Rwanda treaty, followed by the Safety of Rwanda Bill and eventually the Safety of Rwanda Act.
It was an extraordinary amount of legislation. The boats kept coming.
In fact, something worse happened. The asylum system itself started to seize up. The Illegal Migration Act meant asylum claims from people arriving irregularly after 7 March 2023 were put on hold. That isn’t my interpretation of what happened; the Home Office subsequently said so itself.
By June 2023 there were 175,457 people waiting for an initial asylum decision. Look specifically at people arriving by small boat in the year to June 2023 and the figures are extraordinary. There were more than 32,000 asylum applications from small-boat arrivals. Eighty-eight per cent were still waiting for a decision. Just one per cent had received an initial decision.
We had created a chaotic asylum system which was barely processing asylum claims.
Dysfunction is an incentive too
This matters because we normally think about immigration incentives in rather simplistic terms. People talk about benefits, hotel accommodation or whether asylum seekers are allowed to work. But time is an incentive too.
Imagine you are considering paying a smuggler to cross the Channel. Perhaps you have a strong asylum claim. Perhaps you don’t. If you know that Britain will establish who you are, check whether you have previously claimed asylum elsewhere, decide your case reasonably quickly and remove you if your claim fails, that affects the calculation.
But suppose the opposite is true. Britain has lost access to the European fingerprint database, has no effective agreement for returning you to France, and its asylum system is so overwhelmed that there is a good chance nobody will make a decision about your case for years. Even if you eventually lose, you may still spend years in Britain before that happens.
That is an incentive, and no amount of ministerial rhetoric about being “tough” alters it.
The financial consequences of this dysfunction were spectacular. By 2023/24 the Home Office was spending about £4.7 billion on asylum support, around £3 billion of it on hotels. At the peak, more than 56,000 asylum seekers were being accommodated in hotels. Meanwhile the government was pursuing the Rwanda scheme, under which no enforced removal of an asylum seeker to Rwanda ever took place. By June 2024 the Home Office estimated that Rwanda and the associated Illegal Migration Act preparations had cost around £715 million.
This is an important part of the story because performative government isn’t free. We weren’t choosing between an expensive compassionate policy and a cheap tough one. We were spending extraordinary amounts of money on a system which was simultaneously becoming less capable of doing its basic job.
Then something rather boring happened
The asylum system started processing asylum claims again. The retrospective effect of the Illegal Migration Act was removed in July 2024, allowing cases which had been sitting in limbo to progress, and decision-making increased dramatically. In the year to June 2026, around 116,000 people received an initial asylum decision, more than four times the average annual number between 2010 and 2019.
The backlog fell. At its peak in June 2023, 175,000 people were waiting for an initial decision; by June this year it was about 40,000, a fall of 77 per cent. Hotel use fell too. By June 2026 around 16,000 supported asylum seekers were living in hotels, half the number a year earlier and 71 per cent below the 2023 peak. Returns increased as well, with around 41,000 people returned from Britain in the year to June, up 8 per cent.
The outcome of those decisions changed too. In the year to June 2026, 38 per cent of initial asylum decisions resulted in a grant of protection or other leave, down from 48 per cent a year earlier and from a peak of 77 per cent in the year to September 2022. In absolute terms, around 42,000 people were granted protection or other leave and around 73,000 were refused.
That does not mean the system simply became “tougher”. Grant rates are affected by who is applying, what is happening in their countries of origin and which cases are being decided. Syria is an obvious example: its grant rate fell dramatically as the situation there changed. But it does mean that the surge in decision-making was not simply a machine for approving old claims. Large numbers of claims were being refused as well as granted.
None of this makes a particularly exciting political slogan. Process applications, make decisions, grant asylum to people who qualify, refuse people who don’t and remove people who have no right to remain. It sounds suspiciously like administration.
Refugees are supposed to be granted asylum
There is an important qualification here. The five largest nationalities arriving by small boat are now Eritrean, Sudanese, Afghan, Somali and Iranian. Together they account for around two-thirds of small-boat arrivals. These are not, by any reasonable description, unusually safe and peaceful countries. They include people escaping war, authoritarian governments, political and religious persecution and serious human-rights abuses.
That matters because the purpose of an asylum system is not to refuse asylum. A high refusal rate isn’t evidence of success any more than a high grant rate is evidence of failure. The purpose is to distinguish people who need protection from people who don’t. Someone escaping persecution in Eritrea or war in Sudan isn’t evidence that Britain’s border controls have failed simply because we grant them asylum.
The test of a functioning state is whether it can establish the facts and make a decision reasonably quickly: protect the people entitled to protection, refuse those who aren’t, and make both decisions quickly enough for them to mean something.
France matters
The other big change has happened on the other side of the Channel. Britain has increased cooperation with France, with more intelligence sharing, more French enforcement and more attempts to stop boats before they leave. More than 20,000 attempted crossings were prevented during 2025/26 and, during part of this summer, French authorities were stopping around 61 per cent of small-boat events.
Smugglers appear to have responded by putting more people into fewer successful boats. Average occupancy has risen to around 65 people per boat. There is also the UK-France returns agreement. Since August 2025, some people arriving by small boat have been returned directly to France, with Britain accepting an equivalent number through a controlled legal route. By the end of June, 1,087 people had been returned to France under the scheme.
That isn’t enough on its own to explain a fall of nearly 18,000 crossings, but it doesn’t have to be. Deterrence isn’t necessarily about everybody being caught; it is about changing probabilities. The chance of being stopped in France has increased. The possibility of being returned to France has increased. The probability of actually receiving an asylum decision in Britain has increased enormously and, if that decision is negative, the probability of being removed has increased too.
The calculation involved in paying thousands of pounds to a smuggler has changed.
There are other explanations
Weather matters enormously in the Channel. The Home Office itself uses sophisticated measures of potential crossing conditions involving wave height, wind, rainfall and surf conditions. We don’t yet have sufficiently complete comparable weather data to say that differences in Channel conditions played no part in the fall, and almost certainly they played some part in individual months.
That is one reason I was cautious about last August’s figures. But we are no longer comparing one August with another; we are comparing twelve months with twelve months.
Nor does the decline appear simply to be people switching to another obvious illegal route. In the year to June, total detected illegal-route arrivals fell 22 per cent. Inadequately documented arrivals by air fell 24 per cent and people detected after clandestine entry fell 11 per cent. The missing small-boat arrivals haven’t simply reappeared in anything like equivalent numbers in lorries or at airports.
There are other possible explanations. Wars change, nationalities change and smuggling networks adapt. We should therefore be wary of claiming that one government policy produced precisely a 40 per cent reduction. Real life isn’t that tidy. But the direction of travel is becoming difficult to ignore.
The invasion continues
In July, Nigel Farage responded to the decline in small-boat arrivals by declaring: “The invasion continues.” A few days later Reform announced Operation Fortress, apparently the biggest military operation in the English Channel since the Second World War, involving the Royal Navy and Royal Marines.
This was happening while the number of people crossing the Channel was falling substantially.
That is interesting because if your primary concern is controlling Britain’s borders, fewer illegal arrivals should be good news. You can reasonably argue that 26,837 is still far too many, that the government needs to go further, or that the France agreement may not continue to work. But a 40 per cent reduction is surely something you would want to understand.
Instead, the language becomes more martial: invasion, fortress, Royal Navy, Royal Marines, the Second World War.
Why?
Everyone needs to be the hero of their own story
Jonathan Haidt’s The Righteous Mind offers one possible way of thinking about this. Haidt argues that our political beliefs aren’t simply produced by rational calculations about policy; they are rooted in moral instincts too. Among those instincts are loyalty to the group and respect for authority.
There is nothing inherently sinister about wanting to protect your community. It is one of the things that makes communities possible and, in wartime, it can become heroic. People really do risk their lives protecting their families, neighbours and countries from people trying to harm them.
But in peacetime the instinct can curdle into something less attractive. Protecting the community becomes deciding who is entitled to belong to it. Loyalty becomes suspicion of outsiders. The desire to protect becomes a desire to control. Politics can then provide people with an opportunity to imagine themselves in the role of protector.
Small boats are almost perfectly designed for this. Look at the language: Britain is being invaded, our borders have been breached, migrants are flooding across the Channel, the country must be defended. It turns an extremely complicated problem involving refugees, migration, international law, organised crime and public administration into a miniature war.
There are invaders, there are weak politicians who refuse to defend us, and there are plucky defenders of Britain standing on the metaphorical cliffs of Dover, prepared to do what everyone else is too frightened to do.
It is a much more satisfying story than improving the productivity of asylum caseworkers.
The trouble with competent government
Actually controlling a border is incredibly boring. It involves databases, fingerprints, caseworkers, courts, intelligence sharing, police cooperation, international agreements and civil servants making thousands of individual decisions. It means knowing who has arrived, establishing where they came from, deciding whether they have a legal right to remain and doing something about it.
None of this allows anyone to pretend to be Churchill, or fantasise about being Spitfire pilots. But it works a lot better.
Perhaps that exposes something uncomfortable about a particular kind of anti-state politics. Some politicians spend years telling us that government doesn’t work, civil servants are useless, the state is bloated, bureaucracy is the problem and institutions stand in the way. Then they encounter a complicated problem which can only really be solved by a functioning state.
Their answer is another slogan, another law, another crackdown, another enemy and another promise to get tough. When the state becomes less capable as a consequence, the resulting failure becomes further evidence that the state doesn’t work. It is a remarkably effective political feedback loop. It is also extremely expensive.
Toughness isn’t the same as strength
There is a lesson here which goes beyond immigration. Politicians regularly promise to “get tough”. It is attractive politics because toughness can be announced immediately. A new law can be unveiled, a target announced, an enemy identified and a minister dispatched to a television studio. None of these things requires the policy actually to work.
For several years Britain’s asylum policy became progressively tougher in its rhetoric and increasingly dysfunctional in its administration. The legislation became tougher, the speeches became tougher and Rwanda was certainly supposed to be tougher. Meanwhile the asylum backlog grew, hotels filled, the bill increased and the boats kept coming.
What appears finally to have improved matters is almost embarrassingly conventional: process applications, make decisions, protect people fleeing war and persecution who qualify for asylum, refuse those who don’t, remove people who have no right to remain, share information with neighbouring countries, and work with France to stop smugglers putting boats into the Channel.
Most importantly, measure success by whether fewer people actually make the journey rather than by how angry the Home Secretary sounds on television.
One year ago I wrote that one month’s figures didn’t make a trend. A year later, I think we can say that something has changed. Small-boat crossings have fallen by around 40 per cent. We cannot attribute every one of those missing crossings to a particular policy. Weather matters, wars matter, smugglers adapt and migration is complicated.
But there is now a substantial body of evidence pointing in the same direction. The asylum backlog is dramatically smaller, decisions are being made, hotel use has fallen, returns have increased, French interceptions have increased, alternative illegal routes haven’t risen to compensate, and small-boat crossings have fallen.
After years of Rwanda, slogans and successive pieces of immigration legislation, the policy that appears to be working looks suspiciously like competent public administration.
That is worth remembering the next time a politician promises to solve a difficult problem by getting tough. There is a difference between protecting a community and performing the role of its protector.
One is interested in outcomes. The other needs the battle.