Europe's 2015 Migrant Crisis and the Lessons It Taught Every Democracy

A rule that is not enforced is not a rule. It is a suggestion with a stamp on it. In 2015, Europe discovered how many of its most important rules were suggestions.

I want to tell this story carefully and with respect, because the people at the center of it were human beings in real danger, and a good many of them were fleeing a real war. I also want to tell it plainly, because the political lessons are among the clearest of our time, and they apply to every democracy that shares a border, including ours. The story has heroes and it has villains, and the villains were mostly systems: rules that could not hold, institutions that could not decide and signals that no one controlled.

The routes

Start with the geography, because it explains most of what followed. People trying to reach Europe in 2015 came by three main paths. The Central Mediterranean route ran from Libya to Italy, a dangerous crossing used for years. The Eastern Mediterranean route ran from the Turkish coast to the Greek islands, a short but perilous trip in rubber boats. And from Greece, the Western Balkan route carried people overland through North Macedonia, Serbia and then north through Hungary or Croatia, onward to Austria and above all Germany.

In 2015, more than a million people reached Europe by sea, according to the figures that the United Nations refugee agency and the International Organization for Migration published at the time, and most of them crossed from Turkey to Greece. A large share were Syrians fleeing a civil war that had been grinding for years. Others came from Afghanistan and Iraq, and others from Eritrea and elsewhere in Africa and from places where the claim to protection was weaker. That mix matters for the policy lessons, because a system built to protect people fleeing persecution has to be able to tell the cases apart, and the surge overwhelmed the machinery that was supposed to do it.

The rules that were supposed to hold

Two sets of rules governed the situation, and both were put to a test they could not pass.

The first is the Schengen system, under which most EU states and a few others had abolished passport checks at their internal borders. The bargain was explicit: if you can move freely inside, the outside has to be guarded. Control of the external frontier was, by design, the job of the states on the edge, such as Greece and Italy.

The second is the Dublin Regulation, in its third version from 2013. Its core rule is that an asylum seeker's claim is to be examined by the member state responsible, and the main criterion in practice is the country where the person first entered the EU. The idea was to prevent "asylum shopping" and to ensure that someone is responsible for every claim. Read it again, because the logic matters: the frontline countries carry the load. A system that assigns the burden to the states with the longest coastlines and the weakest finances, while the richer states in the center enjoy the free movement, is a system with a built-in point of failure.

The failure came fast. Greece, in the middle of its own financial crisis, could not register or house the people arriving in huge numbers. Many were simply waved onward. Italy had the same problem on a smaller scale. People walked north through the Balkans, and the countries along the route had no capacity or wish to keep them. The Dublin rule said claims belong to the first country of entry. In practice, the first country of entry was not keeping them, and everyone knew it.

The decision in Berlin

The pivotal political act came from Germany. In late August 2015, the German federal migration office, the BAMF, signaled that it would not apply Dublin transfers back to the country of first entry for Syrians. In effect, Germany would take responsibility for examining their claims. On August 31, Chancellor Angela Merkel told reporters, in a phrase that became a slogan, that Germany could manage the task. Then, in the first days of September, thousands of people stranded in Budapest and trying to continue to Germany were allowed to travel on, through Hungary and Austria, in a decision taken by the German and Austrian governments.

I will say what I think the German government was trying to do. It saw a humanitarian emergency, a large number of people in desperate conditions at a railway station and on highways, and it decided to act to relieve it. That is a defensible human impulse, and I do not question the sincerity. Shortly before, the image of a drowned Syrian toddler washed up on a Turkish beach had shaken Europe, and the pressure to act was immense. The question for the policy analyst is not whether the impulse was decent but what the decision did as a signal.

A signal travels faster than a decision. Whatever the legal fine print said, the message that went out along the routes, through phones and smuggling networks and family chats, was that Germany was open. People who had been undecided decided. The flow, already large, grew, and those who profited most reliably from the growth were the smugglers, a point that comes up again in How a Smuggling Network Prices a Border Crossing and Who Gets Paid. Germany's authorities first counted about 1.1 million arrivals in their registration system for 2015 and later revised that to about 890,000, while formal first asylum applications that year were about 440,000. The gap between those numbers is itself a sign of how overwhelmed the registration machinery was, and I would ask you to check the official statistics.

The strain, and the politics

It would be dishonest to describe the strain only in terms of cost, though the cost was real. Housing, schooling, language courses, security screening and the processing of claims all fell on local authorities and volunteers, and in many places they rose to the challenge. German civil society did remarkable things in the autumn of 2015. Ordinary people brought food and clothes, opened their homes and taught the language. It is part of the record, and the right is not obliged to pretend otherwise.

But the strain showed up in politics as well. Within the EU, the disagreement was sharp. A proposal in September 2015 to relocate asylum seekers among member states by a quota system was opposed by several governments in Central Europe, and the relocation scheme delivered only a small fraction of what was planned. Hungary built a fence on its border with Serbia, and then on its border with Croatia, and was widely criticized, while its government replied that it was doing what the Schengen rules required of a frontier state. Sweden, which had taken in a very large number of asylum seekers relative to its population, introduced border checks in November 2015 and later restricted the permits it issued. Others, including Austria and Denmark, reintroduced checks at internal borders, which meant that Schengen itself, the proudest achievement of the European project, was suspended in parts.

Electoral politics shifted. Anti-immigration parties gained across the continent. In Germany, the Alternative for Germany, a party that had begun as a protest against euro bailouts, reoriented around the migration issue and entered the Bundestag in the 2017 election. I note the fact without endorsing any party. A political scientist would observe that voters were responding to something real: a sense that the government had made a large decision without asking them, and that rules they had been told were binding had turned out to be optional.

The EU-Turkey statement

The system's response, when it came, was an agreement with Turkey. On March 18, 2016, the European Union and Turkey issued what is called the EU-Turkey statement. Its main terms were these. Irregular migrants arriving in Greece from Turkey after a set date would be returned to Turkey. For every Syrian returned from the Greek islands to Turkey, another Syrian would be resettled from Turkey to the EU. Turkey would take measures to prevent new sea or land routes of irregular migration from opening, and the EU would provide financial support, in the first phase of several billion euros, for the refugees in Turkey and would revive other parts of its relationship with Ankara.

Meanwhile, the Western Balkan route was closed in practice in the spring of 2016 as countries along it restricted passage, and Greece's neighbor to the north stopped letting people through. The combined effect was a steep fall in arrivals by the eastern Mediterranean route in the months that followed.

Whatever you think of the arrangement, and it has been criticized by humanitarian groups and by those who object to depending on a government with its own agenda, the lesson is clear and not comfortable for the people who said enforcement could not work. A policy of returning arrivals to a safe third country, with a legal channel for resettlement alongside it, reduced the flow dramatically and quickly. The reason was the one every enforcement skeptic denies and every smuggler understands: the incentive to make the crossing depends on what happens to you when you arrive. This is the same logic behind the American rules discussed in Remain in Mexico and Safe Third Country Rules, How Each Is Supposed to Work.

What Europe's later years showed

The aftermath is a lesson in itself. Several governments tightened. Germany limited family reunification for certain categories of protection and put new rules and safe-country designations into law. Denmark moved toward temporary protection. Sweden, which had once been a model of liberal asylum policy, turned sharply in the other direction. The EU spent years negotiating a reform of its asylum rules, and a new Pact on Migration and Asylum was adopted in 2024, after nearly a decade of argument, with tighter screening at the border and a mechanism for sharing responsibility among members.

The contrast with 2022 is worth noting, because it shows what Europe can do with consent and clarity. When Russia invaded Ukraine, the EU activated a temporary protection directive that had been written years earlier and never used. Millions of Ukrainians were given the right to stay, work and send their children to school for a defined period without going through individual asylum procedures. Most were women and children fleeing a neighboring war, the countries receiving them were neighbors, and the framework was agreed in advance. It worked far more smoothly than 2015, which tells you something about the difference between a pre-agreed, time-limited, rule-bound response and one improvised in the middle of a crisis.

The lessons

Let me draw them in numbered form, because they are the point.

One. Signals are policy. A government that says, or seems to say, that anyone who arrives will be admitted has changed the incentives of everyone who is thinking about arriving. The decision to waive Dublin transfers was made for compelling humanitarian reasons, and it also set a magnet. Good intentions do not neutralize incentives, and a policy has to be judged by what it induces as well as by what it intends.

Two. A shared border needs a shared rule and a shared enforcer. The Schengen-Dublin design placed the duty to guard the common frontier on a few states with no means to do it, and left the center free to criticize the results. A structure in which one party carries the burden and another gets the benefit will not hold in a crisis. I think of the Articles of Confederation here. In Federalist No. 42, Madison argued that a power over commerce and foreigners had to belong to the national government, because states acting separately on a shared problem would create friction and, worse, free-riding. Europe in 2015 was a confederation trying to run a common border. The American lesson is that immigration enforcement is a federal job under the Constitution, and a federal government that declines to do it leaves the states to improvise, as the dispute covered in Does Mass Illegal Entry Count as an Invasion Under the Constitution shows.

Three. Humanity and order are allies. The people who suffered most in 2015 were those on the water and in the hands of smugglers, and the policy that reduced the deaths was one that reduced the incentive to attempt the crossing. The arguments of A Secure Border Is the Humane Border, and Here Is the Moral Case apply here with full force. An orderly system can offer a legal path and protect the vulnerable more reliably than a chaotic one.

Four. Democratic consent is not optional. The political backlash was predictable. When a government makes a decision of this scale without a vote, and the decision turns out to have consequences the public did not accept, voters punish it, and they sometimes reach for movements that the establishment finds alarming. A policy that is humane in conception and unaccountable in execution is a recipe for exactly that outcome.

Five. Integration is not automatic. Newcomers who have an immediate path to language learning and work do better, and those who spend years in limbo do worse. Any country that admits people owes them both a clear process and a real chance to join the society. That is as true here as in Germany.

The other side's best case

The fair counter-argument deserves its due. It says: the 2015 arrivals were largely refugees from a real war, and Europe's moral duty under the 1951 Convention was to receive them. Germany's decision upheld the principle that a rich society does not turn away people fleeing barbarism, and the long-run results, many of whom learned the language, found work and built lives, vindicate it. Border closures and deals with Turkey simply moved the suffering out of sight.

I take the first half of that seriously. The duty to the persecuted is real, and it is rooted in the Judeo-Christian tradition that conservatives claim. But the duty is a duty to people who actually qualify, assessed through a process, and a mass movement in which people with strong claims and people with weak ones arrive together cannot be sorted when the system is overwhelmed. The Turkey arrangement did not end the right to seek protection. It redirected it to a process, and it did so while maintaining a resettlement route. Whether that balance was struck perfectly is debatable. That an orderly process serves refugees better than a stampede is not.

The case in brief

1. In 2015, more than a million people reached Europe by sea, mostly through Greece, and the Dublin and Schengen rules, which placed the burden on the frontline states, could not hold.
2. Germany's decision to suspend Dublin transfers for Syrians was humane in intent and, as a signal, it increased the flow that followed.
3. The EU-Turkey statement of March 18, 2016, which paired returns to Turkey with resettlement, sharply reduced arrivals by the eastern Mediterranean route.
4. The lessons for every democracy are that signals are policy, a shared border needs a shared enforcer, order and humanity are allies, and consent matters.
5. The 2022 response to Ukraine shows what a pre-agreed, time-limited, rule-bound framework can do.

What you can do: read Article I, Section 8, Clause 4 of the Constitution, then ask your representative, through the page for your representatives, where Congress has defined by statute how asylum seekers are to be processed, and whether the rules would hold in a surge.