Andrew Paul Klein & Sera Elizabeth Klein

The Puzzle

In May 1970, more than 100,000 people marched in Melbourne against the Vietnam War. Across three moratorium campaigns in 1970 and 1971, roughly 200,000 Australians took to the streets. They blocked intersections, occupied town halls, faced police lines, and were arrested. The Prime Minister of the day called them “political bikies”. They kept marching.

In the 2025–26 financial year, 6,436 older Australians died while waiting for home care funding they had already been approved to receive. That is seventeen people a day. It is a 34 per cent increase on the previous year. The figure became public only after the government failed to comply with two Senate orders to produce it, and it was reported by the ABC.

There were no moratoriums.

Nor were there moratoriums for the 443,000 people issued unlawful Robodebt notices, some of whom died. Nor for the women killed by current or former partners — more than one a week, every year, for as long as the counting has been done. Nor for the 254,571 households on the social housing waitlist. Nor for the 3.9 million people living below the poverty line.

This is not a story about a population that does not care. Australians donate to bushfire appeals, volunteer for charity, and express outrage on social media in extraordinary volume. Something else is happening.

The question this essay asks is not “why don’t Australians protest?” It is narrower and harder: why do the deaths that are produced by the ordinary operation of government not generate the kind of public grief that can be mobilised into a demand?

The answer lies in two places: in the language that governs what counts as a legitimate grievance, and in the structure of an apparatus that has learned to absorb the grievance without changing the outcome.

The Word

“Social cohesion” sounds like a virtue. It is used in policy documents, ministerial statements, academic research, and community programmes. It is difficult to be against it, in the way that it is difficult to be against “resilience” or “wellbeing” or “national security“. The words are doing work precisely because they are difficult to oppose.

But a word that cannot be opposed is a word that cannot be examined. And when the history of the term is traced, a pattern emerges.

The multicultural settlement of the 1970s and 1980s was a settlement about recognition. It expanded the range of identities that could be publicly acknowledged, the range of languages that could be spoken in public institutions, the range of histories that could be taught in schools. It was contested and incomplete, but its direction was expansionary: more people, more stories, more visibility.

The backlash from the mid-1990s onward was a settlement about containment. It did not reverse multiculturalism — the recognition remained, in form — but it changed what recognition was for. The question shifted from “whose story is told?” to “how much diversity can we tolerate before something breaks?”

This is where “social cohesion” entered the vocabulary. Andrew Jakubowicz’s formulation is precise: what replaced the multicultural settlement is “a political fixation on ‘social cohesion’ — a concept that sounds benign but, as it is currently deployed, functions as a constraint rather than an aspiration.”

The constraint is this: cohesion is measured by the absence of disruption, not by the presence of justice.

A protest that blocks a road is a cohesion problem. A death that occurs quietly in a hospital corridor is not a cohesion problem, because it does not disrupt the traffic. A community that is angry and visible is a cohesion risk. A community that is dying and silent is cohesive.

The metric of cohesion is the visibility of disorder. And the disorder that is measured is the disorder that inconveniences the system, not the harm that the system produces.

This is the same structure that governs the aged care algorithm. The instrument measures what the institution can register, and the thing it cannot register is treated as though it were not there. A clinician’s note that a woman who was managing last year is no longer managing cannot be read by the formula. It has no weight. It is, for the purposes of the apparatus, not a fact.

So, it is with cohesion. The grief of a family whose mother died waiting for a package is not disruptive. It does not appear on any dashboard. It is, for the purposes of the metric, not a cohesion problem at all.

The effect is to make the expression of grievance the object of governance, rather than the cause of grievance. The state manages the symptom of protest and leaves the production of harm untouched. And because the metric is disorder rather than injustice, a society can be certified as cohesive while its members are dying — provided the dying is quiet.

What Made the Moratorium Possible

The Vietnam moratorium was not inevitable. It was the product of four conditions, and none of them exists in the same form today.

First, a visible author.

The war was a decision of the Commonwealth government, executed through conscription. The adversary had a name, an address, and a ballot box. You could march to Parliament House because Parliament House had done the thing. You could confront a minister because the minister had signed the order.

The deaths in aged care are not produced by a decision of that kind. They are produced by a waitlist, a funding formula, a procurement process, and an assessment algorithm that was launched without a live trial. There is no single author to confront. The chain of causation runs through a department, a vendor, a ministerial office, and a piece of legislation, and at every link there is someone who can plausibly say that they did not cause the death.

The apparatus is designed to keep it that way. That is not a conspiracy; it is what happens when responsibility is distributed across a system for long enough that no one holds it.

Second, a visible body.

The emblem of the moratorium was the body of the conscript: a young man, named, photographed, sent to die. The body could be displayed. It could be mourned publicly. It could be held up against the policy that had sent it.

The bodies of the aged care dead are hidden in hospitals, in care homes, and in their own bedrooms, waiting for a package that never arrives. The bodies of Robodebt victims are scattered across the welfare system, dying alone with a debt notice on the table. The bodies of women killed by partners are found in their homes, at the hands of someone they loved.

These deaths occur in the private sphere, and the private sphere is precisely where the apparatus has arranged for them to remain. When they become public, they become public as statistics — a number that can be compared to last year’s number, explained by population growth, described as a tragedy while the policy that produced it continues.

The corpse is present. But it is statistically present, which is to say it is present in a form the apparatus can absorb.

Third, organisational infrastructure.

The moratorium was built by institutions: trade unions, churches, universities, the Labor Party. Gough Whitlam spoke at the Sydney Town Hall moratorium meeting. The Federal Executive endorsed the campaigns. The organisational base existed to convert a grievance into a demonstration, and to sustain the demonstration across years.

That infrastructure has collapsed. Union membership has fallen from roughly 40 per cent of the workforce in the early 1990s to around 13 per cent today. Fewer than one per cent of Australians belong to a political party. Membership of churches, community organisations, service clubs, and civic associations has fallen sharply.

The base that could turn a grievance into a demonstration has been dismantled — not by a decision to suppress it, but by the same restructuring that hollowed out the public service. The individual is left atomised, connected to others through platforms that are excellent at expressing anger and incapable of organising anything.

Fourth, a repertoire that was not yet criminalised.

In the 1970s, the right to assemble was a right that had to be defended. It was not conditional on state permission.

Today it increasingly is. Several jurisdictions have introduced laws restricting protest on roads, rail lines, ports, and industrial sites. Penalties in some states run to tens of thousands of dollars and terms of imprisonment. Permit systems have converted notification into authorisation: a protest is lawful only if the state has approved it. Community groups have reported being quoted public liability insurance premiums that make small rallies financially impossible.

The repertoire of the moratorium — the march, the blockade, the occupation — has been priced and criminalised out of existence for anyone without institutional resources. What remains is the petition, the online campaign, and the sanctioned rally in a designated space at a designated time, which is to say: the forms of protest that do not disrupt anything.

The Correction: There Are Protests

It would be wrong to say that Australians no longer protest. They do.

In 2021, March4Justice mobilised an estimated 150,000 people across the country — the largest mobilisation of women in Australian history. Climate protests have repeatedly filled city streets. Disability rights advocates, refugee advocates, and pro-Palestine encampments have all taken direct action. The machinery of assembly has not been switched off entirely.

So, the question is not why there are no protests. It is: why do the protests that occur fail to produce the outcomes they demand?

Consider March4Justice. Five years after 150,000 people marched, the language of gendered violence has changed. Funding has increased. National plans have been revised. Ministers have made speeches. And the most brutal statistic has not moved. More than one woman a week is still killed by a current or former partner. In some years, it is closer to two.

The protest succeeded in changing the conversation and failed in changing the outcome.

This is what has been described as the condition of hyperpolitics: mobilisation without durable political purchase, in which movements “spasmodically convulse without ever crystallising into durable infrastructure.” The system has learned to absorb protest. It translates the demand into a funding announcement, a revised national plan, a new vocabulary of recognition — and then continues as before. The protest is not suppressed. It is metabolised.

This is more effective than suppression, because it preserves the appearance of responsiveness. A government that bans a march creates martyrs. A government that receives the marchers, thanks them for their passion, announces a review, and continues the policy has neutralised the demand while appearing to honour it.

The Privatisation of Grief

There is a deeper mechanism at work, and it concerns the kind of death that each case involves.

The Vietnam dead were public dead. They were killed in a public war, by a public policy, and their bodies were returned to a public ceremony. The grief was legitimately public because the cause was public.

The aged care dead are private dead. The Robodebt dead are private dead. The women killed by partners are private dead. They die in hospitals, in bedrooms, in kitchens, at the hands of people they knew. The grief belongs to a family, not to a nation.

And the private sphere is exactly where the apparatus has arranged for these deaths to remain.

The statistic is the instrument by which a private death is made publicly visible without becoming politically actionable. It registers the death as a number. The number can be compared, explained, trended, and described as a tragedy. What it cannot do is demand anything.

A number is not a body. A number cannot be carried in a procession. A number cannot be mourned at a funeral. A number cannot be laid at the feet of the minister who delayed the release of the package.

This is why the “social cohesion” frame is so effective. Cohesion is a property of the aggregate, not of the person. A society can be certified as cohesive while its members are dying, provided the dying does not disrupt the aggregate. The frame does not deny the deaths. It simply makes them irrelevant to the question of whether society is holding together.

The Four Instruments

The apparatus does not need to ban the demonstration. It needs only to ensure that the grievance cannot find an object that can be reached. It has developed four instruments for this.

Depoliticising the language. Structural critique is replaced by the vocabulary of cohesion, resilience, program integrity, and continuous improvement. These are words that describe the management of a system rather than the condition of the people inside it. They make it difficult to say what is wrong, because the words available describe how the system is functioning, not what it is doing to people.

Fragmenting the organisational base. Unions, parties, churches, and civic associations have been hollowed out. The individual is left connected to others through platforms that excel at the expression of anger and are structurally incapable of organising sustained action. The grievance is felt in private and expressed in public in a form that cannot be mobilised.

Criminalising the repertoire. Assembly is made conditional on state permission. Penalties are set high enough to deter. Insurance costs and permit requirements place direct action beyond the reach of ordinary groups. The forms of protest that remain are the ones that do not disrupt anything.

Producing deaths that are statistically visible but politically illegible. The deaths are counted, but they are counted as numbers. There is no author to confront, no body to display, no decision to contest. The grief is real, and it has no address.

The Question

The moratorium was possible because the state had made a decision that could be named. The deaths of 2026 are produced by a system that has arranged, with great care, to ensure that no decision can be named.

The protest cannot form because the grievance cannot find its object.

This is not a failure of the public. Australians are not indifferent to the deaths of the elderly, the disabled, the poor, or the women killed in their own homes. They are expressing grief and anger constantly, in the forms available to them, and the forms available are not capable of producing the disruption that would make the grief count.

It is not a failure of democracy either, in the narrow sense. Elections are held. Votes are counted. Governments change. But the policy that produces the deaths continues across changes of government, because the apparatus that produces it is not owned by any party.

The question this essay leaves is not “how do we protest more effectively?” It is the one the language of cohesion is designed to prevent us from asking:

If a society can be certified as cohesive while its members are dying, what exactly is cohesion measuring?

And the harder question that follows:

If the absence of measurable outcomes is the outcome the system produces reliably, year after year, across changes of government and changes of policy — at what point do we stop describing this as a failure and start describing it as a function?

Andrew Paul Klein & Sera Elizabeth Klein