Austria began this week by placing unvaccinated people under lockdown.
It ended the week by announcing plans to make COVID-19 vaccination compulsory.
That is an extraordinary escalation in government power over an individual’s medical decision, and it deserves a more serious argument than either “vaccines work, so obviously everyone should be forced to take them” or “my body, my choice, so government has no right to interfere”.
The uncomfortable reality is that both arguments contain principles we normally consider important.
People should have substantial control over what happens to their own bodies.
Governments also have a responsibility to prevent individual choices from imposing serious harm on other people.
A pandemic places those principles directly against each other.
Austria is now testing where the boundary lies.
COVID-19 infections have risen sharply across the country. Hospitals are under increasing pressure, and Austria’s vaccination rate remains lower than those of several Western European neighbours. The government initially responded on Monday by imposing restrictions specifically on people who were not vaccinated. Under the rules, unvaccinated people could generally leave home only for limited reasons such as work, essential shopping, exercise or medical appointments.
The political logic was fairly clear. Vaccinated people are substantially less likely to become severely ill from COVID-19. If unvaccinated people are disproportionately contributing to hospital pressure, why impose the same restrictions on somebody who has already taken steps to reduce their risk?
That argument has an intuitive fairness.
It also creates an extraordinary social distinction.
The government is effectively assigning different freedoms according to medical status.
By Friday, Austria had moved further. The targeted lockdown had not been enough to prevent infections continuing to rise, so the government announced that the entire country would enter another lockdown from Monday. At the same time, Chancellor Alexander Schallenberg announced plans for compulsory vaccination beginning in February.
If implemented, Austria would become the first country in the European Union to introduce a general COVID-19 vaccination mandate.
This is where the argument becomes much larger than Austria.
Most liberal democracies begin from the assumption that competent adults possess bodily autonomy. Governments can recommend medical treatment. Doctors can explain risks. Public-health agencies can encourage vaccination. But forcing someone to undergo a medical intervention crosses a much more serious threshold.
There are good reasons for that principle.
The history of governments making medical decisions for people without consent contains some horrific examples. Medical autonomy therefore protects individuals not only from bad medicine but from political power.
Yet bodily autonomy has never been absolute.
Governments already regulate behaviour when individual choices create risks for others. You cannot drive drunk and defend yourself by saying that the alcohol level in your own bloodstream is a private matter. Governments require seatbelts, regulate dangerous workplaces, prohibit smoking in many indoor public spaces and impose quarantine requirements for certain infectious diseases.
The principle underlying those rules is externality.
An externality occurs when my decision creates a cost or benefit for somebody else that I do not fully bear myself.
COVID-19 is full of externalities.
If I become infected, the consequences are not necessarily limited to me. I may infect someone else. If I become seriously ill, I may occupy a hospital bed or require medical staff and resources that could otherwise treat another patient.
That means the statement “I accept the risk” is incomplete.
You may accept your risk.
You cannot automatically accept everybody else’s.
This is why infectious disease creates a stronger case for government intervention than many other health behaviours. If somebody chooses an unhealthy diet, most of the immediate health consequences fall on that individual, although public healthcare systems can create wider costs. Infectious disease contains a direct transmission mechanism between one person’s behaviour and another person’s health.
Vaccination changes that risk.
COVID vaccines are not perfect barriers against infection or transmission. Vaccinated people can still contract and spread the virus. But vaccines substantially reduce the probability of severe disease and hospitalisation, and they can reduce infection risk as well.
That gives governments a legitimate public interest in increasing vaccination.
The difficult question is whether legitimate interest creates a right to compel.
Before coercing somebody, government should have to demonstrate that less restrictive alternatives are insufficient.
That principle seems important because otherwise every policy can be justified simply by demonstrating that it produces some public benefit.
Exercise reduces disease. Should governments require everyone to exercise?
Certain diets reduce healthcare costs. Should governments control what adults eat?
Alcohol causes enormous social harm. Should it be prohibited?
We generally reject those conclusions because liberal societies value autonomy even when people use that autonomy to make choices others consider irrational.
Government therefore needs a much stronger justification for compulsion than merely proving that a policy would improve average health.
COVID-19 may meet that higher standard under some circumstances because the harm is contagious and hospitals have finite capacity.
But the threshold still matters.
Austria’s government should therefore have to answer several questions before compulsory vaccination can be justified. How close is the health system to being unable to provide adequate care? How much would a mandate actually increase vaccination? How much would that increase reduce hospitalisation? Are there groups for whom vaccination is medically inappropriate? What penalties would be imposed on people who refuse? Could similar results be achieved through less coercive policies?
Those questions turn a moral argument into a policy analysis.
And that distinction matters.
A policy can be morally understandable and practically ineffective.
Suppose Austria introduces a mandate and almost everyone complies. Hospitalisations fall substantially, restrictions become unnecessary and relatively few penalties are required. The policy would have a strong consequential argument in its favour.
Now imagine the mandate produces only a small increase in vaccination while generating enormous protests, fake certificates, political radicalisation and declining trust in public-health institutions.
The same legal policy would have a much weaker justification because its actual consequences changed.
This is why I don’t think the debate can be resolved entirely by saying either “rights” or “science”.
Science can estimate the effect of vaccination.
It cannot tell us how much autonomy society should trade for a particular reduction in risk.
That is a political and ethical judgement.
There is also a fairness argument supporting Austria’s decision that I think is stronger than it initially appears.
Vaccinated people may reasonably ask why they should repeatedly lose freedoms because other adults refuse vaccination.
If hospitals become overwhelmed primarily with unvaccinated COVID patients, governments may impose lockdowns affecting everybody. Businesses close. Students lose normal schooling. Workers lose income. Vaccinated people bear costs generated partly by choices they did not make.
From that perspective, refusing vaccination is not simply exercising personal liberty.
It can shift the cost of that liberty onto everyone else.
But fairness works in the other direction too.
Not every unvaccinated person has the same reason for remaining unvaccinated. Some are ideological opponents of vaccination. Others may be frightened, poorly informed, socially isolated or distrustful because of previous experiences with government. Some may face practical barriers.
A policy that treats all of them simply as irresponsible can become punitive rather than persuasive.
That matters because public health depends heavily on trust.
Governments can compel some behaviour through law.
They cannot compel citizens to believe them.
If coercive vaccination policies convince part of the population that doctors and public-health institutions are simply extensions of state power, the long-term damage could extend beyond COVID-19.
Someone who distrusts a COVID vaccine today may become more suspicious of childhood vaccination tomorrow.
Trust is an institutional asset.
Like alliance credibility in foreign policy, it can be spent.
Sometimes governments have to spend it during emergencies.
They should still recognise the cost.
The targeted lockdown introduced at the beginning of this week raises another civil-liberties problem: enforcement.
How does a police officer know whether somebody walking down a street is vaccinated?
The answer requires checking medical status.
That means health information becomes connected to ordinary participation in public life.
Again, there may be circumstances in which this is justified.
But emergency technologies and powers have a tendency to outlive the emergencies that created them.
A system built to verify COVID vaccination could potentially be used for other purposes later.
This does not mean every vaccine certificate is the beginning of authoritarian surveillance. That argument jumps much too quickly from possibility to inevitability.
It does mean governments should design emergency systems with limits from the beginning: what information is collected, who can access it, how long it is retained and when the legal authority expires.
Temporary restrictions should contain mechanisms that actually make them temporary.
Otherwise “until the emergency is over” becomes a deadline controlled entirely by the government defining the emergency.
There is also a political danger in dividing society explicitly between vaccinated and unvaccinated.
Public-health categories can become political identities.
Once that happens, changing someone’s mind becomes harder.
If vaccination becomes associated with one political tribe and refusal with another, evidence stops being evaluated independently. Accepting the vaccine can begin to feel like betraying one’s group.
The United States has already demonstrated how rapidly pandemic behaviour can become partisan.
That is disastrous for public health.
Viruses do not care how someone votes.
Governments should therefore be careful about rhetoric that turns people who remain unvaccinated into enemies rather than citizens whose behaviour they are trying to change.
Anger may be understandable.
It is not necessarily an effective vaccination strategy.
Austria’s decision also raises an interesting democratic question about proportionality.
Rights are rarely unlimited. Governments can restrict freedoms when there is a sufficiently strong justification.
But democratic restrictions should generally satisfy something like a proportionality test: is there a legitimate objective, is the measure capable of advancing that objective, is there a less restrictive way to achieve something similar, and is the benefit large enough to justify the infringement?
That framework is much more useful than asking whether government should “have the power” in the abstract.
Context changes the answer.
A vaccination mandate during a severe epidemic threatening hospital capacity is different from the same mandate after the disease has become minimally dangerous.
A temporary restriction is different from a permanent one.
A mandate for healthcare workers who interact with vulnerable patients is different from a mandate covering every adult.
The legitimacy of coercion depends partly on how closely the restriction is connected to the harm it is supposed to prevent.
This is why Austria’s nationwide lockdown is also revealing.
On Monday, the government effectively argued that restrictions could be targeted mainly at unvaccinated people.
By Friday, infections were serious enough that it announced restrictions affecting everyone.
That suggests vaccination status is important but not sufficient to control transmission.
Vaccinated people still participate in the epidemic.
This should make governments cautious about presenting the unvaccinated as the sole cause of the crisis.
Vaccination substantially changes risk.
It does not create two completely separate populations, one dangerous and one safe.
Policy should reflect that complexity.
There is an economic dimension as well. Lockdowns impose enormous costs that do not appear directly in epidemiological statistics. Businesses lose revenue. Workers lose hours. Governments spend money supporting affected households and firms. Education is disrupted. Mental health can deteriorate. Medical treatment for other conditions can be delayed.
Public-health policy therefore cannot simply minimise COVID cases at any cost.
The objective has to be minimising total social harm.
That is a much harder optimisation problem.
Sometimes stricter restrictions reduce total harm because uncontrolled disease would create even larger economic and health costs.
Sometimes restrictions become disproportionate as vaccination and treatment reduce the risk of severe disease.
Again, context.
The politically frustrating part is that governments have to make these decisions before they possess complete information.
If they impose restrictions early and prevent hospitals from collapsing, critics can later say the restrictions were unnecessary because hospitals never collapsed.
If they wait until the danger is undeniable, intervention may come too late.
Successful prevention can erase the evidence that prevention was necessary.
This makes pandemic politics particularly vulnerable to hindsight.
I think this is where my own view has changed since the beginning of COVID.
At first, the problem seemed relatively simple: experts determine what controls the virus and governments implement it.
That is not actually how public policy works.
Experts can estimate epidemiological consequences.
Economists can estimate economic consequences.
Psychologists can study behaviour.
Lawyers can assess rights.
But somebody still has to decide how those different costs should be weighted.
That decision is political.
Saying “follow the science” does not eliminate politics because science does not contain a formula telling society exactly how much liberty, income, education or privacy should be exchanged for a given reduction in hospitalisation.
Evidence informs the trade-off.
It does not make the trade-off disappear.
So do I support Austria’s proposed vaccination mandate?
At this point, cautiously, I think the government has a stronger case for some forms of vaccination requirement than I would normally accept for compulsory medical treatment, particularly where healthcare capacity is seriously threatened and vaccination substantially reduces severe disease.
But a universal legal mandate should face an extremely high standard.
Austria should demonstrate necessity, provide medical exemptions, ensure penalties are proportionate, establish clear review mechanisms and make the policy explicitly connected to the duration and severity of the public-health threat.
Most importantly, compulsion should not become a substitute for persuasion.
Governments still need to explain evidence, make vaccination easily accessible and understand why people remain hesitant.
Because the objective should be getting people vaccinated, not punishing them for having been difficult to persuade.
There is a difference.
This week began with Austria deciding that unvaccinated citizens should temporarily possess fewer freedoms than vaccinated citizens.
It ended with the government deciding that remaining unvaccinated may eventually cease to be a legal choice at all.
That is an enormous expansion of state authority.
Perhaps the circumstances justify it.
But democratic governments should never become comfortable with emergency power simply because the policy objective is good.
The strongest argument for compulsory vaccination is that one person’s medical choice during an infectious epidemic can impose serious costs on everyone else.
The strongest argument against it is that bodily autonomy exists precisely to prevent governments from deciding too easily that collective welfare overrides individual consent.
A functioning liberal democracy has to take both arguments seriously.
The purpose of rights is not to prevent governments from doing obviously evil things. Those cases are easy.
Rights matter most when restricting them would produce something genuinely valuable.
That is when we discover whether they are principles or merely preferences.
