For most of the past three decades, an asylum seeker arriving in Germany did not get the same health care as everyone else. For a waiting period, which a 1997 reform stretched from 12 to 36 months, entitlements were restricted to acute illness and pain, usually with a local welfare office acting as gatekeeper. Only after that did people move into regular access.
So Germany effectively ran two systems side by side for the same population. Which one cost more per person?
The restricted one. A quasi-experimental study of national data from 1994 to 2013, published by Kayvan Bozorgmehr and Oliver Razum in PLOS ONE in 2015, found annual per capita spending about 40 percent higher in the restricted-access group, roughly 376 euros more per person per year. That finding should have ended a policy argument that is still going strong a decade later.
The argument that refuses to die
The case for restricting health care to refugees and migrants usually rests on cost. Full access, the reasoning goes, will attract people, strain budgets and crowd out citizens. So entitlements are narrowed to “acute” or “emergency” care, gatekeeping is added, and paperwork is layered on top.
Bozorgmehr and Razum found no evidence that giving asylum seekers regular access increased spending. To be fair to the data, differences in age and type of accommodation between the groups explained a large share of the gap. But the 1997 reform that lengthened the restricted period widened the spending gap further, which is hard to square with the idea that restriction saves money. The likely mechanism is not mysterious. When people cannot see a general practitioner early, conditions that could have been handled cheaply in primary care arrive later, sicker and more expensive, often through the emergency department.
That is the uncomfortable part for both sides of the political debate. Restriction does not just fail on humanitarian grounds. It fails on the fiscal grounds used to justify it.
Scale changes the stakes
This is not a niche question. UNHCR’s Global Trends report counted 123.2 million people forcibly displaced worldwide at the end of 2024, about one in every 67 people on earth. Most of them are not in Europe. They are internally displaced within their own countries or hosted by neighbouring states that are themselves under strain.
For every one of those host health systems, the same design choice comes up. Do you fold newcomers into the existing system, or do you build a separate, thinner track for them?
The separate track is tempting because it looks controllable. It has its own budget line and its own rules. But separate tracks tend to have three predictable properties.
- They are administratively expensive, because someone has to decide eligibility case by case.
- They push care downstream, toward emergency settings where treatment costs more.
- They generate confusion among clinicians, who often do not know what a patient is entitled to and default to caution.
The real barrier is often not the law
Here is where the conventional framing gets it wrong. Debates focus on legal entitlement: what does the statute say a refugee or migrant can receive? But the WHO’s first World report on the health of refugees and migrants, published in 2022, lists a much longer set of barriers: precarious legal status, discrimination, language, administrative and financial obstacles, lack of information about entitlements, and fear of detention and deportation.
Notice how few of those are about what the law formally allows.
A migrant may be entitled to antenatal care and still not come, because she does not know she is entitled, because the registration form is in a language she cannot read, or because she worries that her details will be passed to immigration authorities. On paper, access exists. In practice, the clinic door is closed.
This is why expanding entitlements alone often disappoints. Health ministries announce inclusion, and utilisation barely moves. The entitlement was never the binding constraint.
Fear is a clinical risk factor
The fear point deserves more attention than it gets from health professionals, because it is one we can do something about.
When people believe that seeking care will expose them to enforcement, they delay. They skip vaccinations for their children. They avoid testing for infectious diseases. They present late with conditions that were manageable. None of this stays inside migrant communities. Infectious disease does not check immigration status, and neither do the costs of late presentation.
Some health systems have tried to address this with explicit “firewalls”: policies that separate health data from immigration enforcement, so patients can be confident their records will not be shared for that purpose. The logic is public health logic, not immigration policy. You cannot run disease surveillance or vaccination on a population that is avoiding you.
The trade-off is real and worth naming. Governments want data and control. Health systems need trust. When those collide, health systems usually lose the argument, and the public ends up less safe as a result.
What actually moves the needle
If the evidence points anywhere, it points toward boring integration rather than dramatic new programmes.
- One system, not two. Give newcomers the same card and the same front door, as the German data suggest, rather than a parallel gatekept track.
- Make entitlements legible. Clear, translated information, delivered through community organisations that people already trust, does more than a revised statute nobody has read.
- Fund interpreters as clinical infrastructure. Language support is not a courtesy. It determines whether consent, diagnosis and medication instructions are real.
- Protect health data. Say publicly and credibly what will and will not be shared, and then stick to it.
None of these is glamorous. All of them are cheaper than the downstream costs of exclusion.
Follow one patient through the restricted system
It helps to walk through what restriction looks like from the inside, because the cost does not appear in any single decision.
Imagine a man in his forties who arrived as an asylum seeker a few months ago. He has type 2 diabetes, diagnosed in his home country, and his tablets ran out somewhere on the journey. Under a restricted regime, his condition is chronic, not acute. It sits in a grey zone. To see a doctor, he first needs authorisation from an office that is not staffed by clinicians. He waits. He does not speak the local language well, so the forms take longer. The official, uncertain whether diabetes counts, approves a single visit.
The doctor, also uncertain about what will be reimbursed, prescribes the cheapest option and asks him to return in a month. The next authorisation is delayed. He runs out again. Some weeks later he arrives at an emergency department with dangerously high blood sugar and an infected foot.
Now the system is paying for an emergency admission, possibly surgery, and weeks of inpatient care. Every individual in that chain behaved reasonably under the rules they were given. The rules produced the most expensive outcome available.
That is the pattern the German spending data captures in aggregate. It is not that anyone decided to spend more. It is that restriction moves care from the cheapest point in the system to the most expensive one, and adds an administrative layer on top.
The strongest counterargument
It would be dishonest to pretend the case against open access has no substance. The strongest version goes something like this.
Health systems are already stretched. Waiting lists are long, general practices are full, and staff are exhausted. Adding a large new population with complex needs, including trauma, untreated chronic disease and conditions rarely seen locally, puts real pressure on real services. Citizens who have waited months for an appointment may reasonably ask why newcomers should join the same queue on the same terms. And governments worry, not without reason, that generous entitlements could influence where people choose to seek asylum.
Parts of this hold. Capacity pressure is real, particularly in towns and cities that receive large numbers of arrivals in a short period. Integration into the regular system only works if that system has room, and in many places it does not.
But notice what the counterargument actually supports. It is an argument for adding capacity where newcomers are concentrated, not for restricting their access. Restriction does not reduce the underlying need. It delays it and redirects it into emergency care, which is the part of the system under the most pressure of all. The man with diabetes still ends up in a hospital bed, just later, sicker and at greater cost.
On the question of whether health entitlements drive migration decisions, the evidence is thin, and people fleeing persecution and conflict are generally moving away from danger rather than toward particular benefit packages. Policymakers who design health restrictions as a deterrent are making a bet with patients’ health as the stake, and with limited evidence that it pays off.
Second-order effects that rarely get counted
The direct costs of restriction are only part of the picture. Several indirect effects deserve attention.
Clinician moral injury. Doctors and nurses who are required to ration care by immigration status, or who are unsure whether they are allowed to treat, experience a specific kind of distress. They know what the patient needs and cannot provide it. Over time, that erodes morale and retention in the services that see the most newcomers.
Data blindness. Parallel systems often have parallel records, or none. That makes it harder to track infectious disease, vaccination coverage or chronic disease outcomes among newcomers, precisely the information public health authorities need.
Children. Restrictions aimed at adults frequently catch children too, through missed vaccinations, delayed developmental assessments and untreated mental health problems. Those costs show up years later in schools and social services.
Trust that outlasts policy. Communities remember how they were treated on arrival. A population that learns early that health services are hostile or dangerous may keep avoiding them long after their legal status is settled, which undermines public health efforts for years.
What this means for different decision-makers
For health ministries, the practical test is simple. Before introducing or maintaining any restriction, ask for an estimate of the downstream emergency and inpatient costs it will generate, and of the administrative cost of running the gate. If nobody can produce those figures, the policy is being made blind.
For hospital and clinic leaders, the priority is clarity. Staff should know exactly what newcomers are entitled to, with a named person they can ask when unsure. Ambiguity reliably produces under-treatment, because frontline staff default to caution. Leaders can also make sure interpreter services and trauma-informed care are funded in the services that see the most arrivals.
For funders and humanitarian agencies, the lesson is to avoid building permanent parallel systems. Stand-alone refugee clinics can be vital in an emergency. But if they become the only route to care, they entrench separation and are vulnerable when funding ends. Money spent helping national systems absorb newcomers usually lasts longer.
For clinicians, there is a role in advocacy grounded in what they see. Documenting cases where delayed access led to worse outcomes, anonymised and aggregated, gives policymakers the evidence they are otherwise missing.
What to watch next
A few signals will show whether systems are moving in the right direction.
Watch whether countries receiving large numbers of displaced people integrate them into national health insurance or public systems, or build separate tracks. Watch whether health data is formally protected from immigration enforcement, and whether that protection is communicated to communities in a way they believe. And watch for evaluations that measure not just what newcomers are entitled to, but whether they actually get care, and at what stage of illness.
That last measure is the one that matters. Entitlement is a legal fact. Access is a clinical one.
The takeaway
The political debate treats migrant health access as a trade-off between compassion and cost. The best available evidence suggests that framing is false. Restriction tends to cost more, not less, and the barriers that matter most are often administrative and psychological rather than legal.
Health professionals are well placed to say this plainly, because we see the late presentations, the missed vaccinations and the emergency admissions that restriction produces. That is not an ideological position. It is a clinical observation with a price tag attached.
Where have you seen a gap between what refugees and migrants are entitled to and what they can actually get? Tell us in the comments.
