The Libertarian Answer to "People Are Hurting" -- 2

The previous essay in this series stopped at a deliberately uncomfortable place. It granted the harder case — that the hurt in front of us is not a curated instance doing rhetorical work for someone else's platform, but simply, unmistakably real — and then argued that even real hurt doesn't specify its own remedy. That was the negative half of the answer: what "people are hurting" doesn't license.

This essay owes the positive half. Grant the hurt. Grant that it's real, unselected, and needs relieving now, not after a committee reports back. What does a libertarian actually propose to do about it?

The answer is not "nothing," and it was never "nothing." It's a specific claim about which mechanisms relieve real hurt well, and which only look like they do because the alternative was never seriously tried.

What used to exist, before it was crowded out

Before the modern welfare state, a large share of the working population in Britain and the United States was covered against sickness, injury, and death not by a public bureaucracy but by mutual aid — fraternal societies, friendly societies, ethnic and religious benefit associations, and union-run sick funds, all of them voluntary, member-funded, and locally administered.

The historian David Beito's research on American fraternal societies puts the scale in concrete terms: by the early twentieth century, something on the order of a third of adult American men belonged to at least one fraternal organization providing sickness or death benefits, and Britain's friendly societies covered a comparable share of the working class before the National Insurance Act of 1911. These were not charities in the one-directional sense — a member paid dues, and in return held an enforceable claim on the society's benefit fund when sickness or death hit. It was insurance, run by the people who were themselves the insured, priced and administered by people with direct, local knowledge of the community they served.

That's exactly the knowledge Hayek's argument, from the previous essay, says a national program can't have. A lodge secretary who knew every member's trade, health, and family situation could catch a fraudulent claim or extend informal grace to a member going through a hard patch in ways a distant caseworker, following a national eligibility rule, structurally cannot.

The system wasn't perfect, and it wasn't universal — coverage gaps existed, and they were real. But it existed at meaningful scale, it was accountable to the people who funded it, and it was displaced, not out-competed.

Displacement, not defeat

The displacement is documented, not speculative. Britain's 1911 National Insurance Act and the United States' 1935 Social Security Act didn't simply add a public option alongside the mutual aid sector. In Britain's case, the state initially routed benefits through "approved societies," including the friendly societies themselves, but progressively centralized administration and eligibility rules until the societies' actual function had been hollowed out from underneath their name.

In the American case, Social Security offered a payroll-tax-funded, actuarially opaque, and politically permanent alternative that fraternal sick funds simply could not price-compete against, since one side could set its own price by statute and the other had to earn it from members who now had a mandatory alternative already deducted from their paycheck. Membership in fraternal beneficiary societies collapsed across the following decades, and it didn't recover.

None of this is an argument that the harm caught by these programs went uncaught — some of it plainly was caught, and that matters. It's an argument about what got lost in the substitution: local accountability, member ownership, and the discovery process by which a decentralized system finds and corrects its own failures, replaced by a single, nationally uniform answer to a question — how should this particular person's hurt be relieved — that Hayek's argument says has no single right answer to find.

What the evidence says works now

The modern case for a specific, testable alternative comes from an unlikely place: development economics, and specifically from randomized controlled trials of direct, unconditional cash transfers. GiveDirectly and the researchers who have studied it have produced some of the most rigorously tested welfare interventions in existence, and the finding that keeps replicating is uncomfortable for anyone invested in bureaucratic administration: giving people cash, with no conditions and no case management, performs as well as or better than comparable in-kind and conditional programs on outcomes from consumption to psychological well-being to, in several studies, local economic activity — without the overhead of determining eligibility, verifying compliance, or delivering goods rather than money.

The mechanism isn't mysterious. A hurting person almost always knows their own most urgent need better than a program designed around the median case does, which is Hayek's knowledge problem confirmed empirically rather than merely argued philosophically.

Milton Friedman reached a version of this conclusion by a different route decades earlier, proposing a negative income tax as the least-bad form a government safety net could take if a safety net was going to exist at all — a simple cash floor, phased out gradually with earned income, replacing the tangle of categorical, in-kind programs each with its own eligibility bureaucracy, caseworkers, and perverse marginal incentives.

Friedman's proposal and GiveDirectly's evidence point at the same structural insight from opposite directions: whatever the state's role turns out to be, cash beats category. Money is fungible in a way that food stamps, housing vouchers, and job-training slots are not, and fungibility is precisely what lets local knowledge — the recipient's own — do the allocating instead of a distant administrator's guess.

Answering the scale objection

The standard objection arrives quickly: voluntary charity is nice, but it cannot scale to catch everyone, and mutual aid societies had exactly the coverage gaps a universal program claims to close.

The objection conflates two different mechanisms that this essay has kept separate on purpose. Charity is asymmetric — a gift, with no enforceable claim running the other way, and it is genuinely vulnerable to free-riding and to the fact that a donor's attention is a scarcer resource than a recipient's need. Mutual aid is not charity. It's a contract: dues paid in exchange for an enforceable benefit, priced and administered by the same community bearing the risk, which is exactly the structure that solves the free-rider problem charity alone cannot — nobody free-rides on an insurance pool they didn't pay into, because they have no claim on it.

The historical coverage gaps in the fraternal system are a real critique of how far that model had scaled by 1910, not a critique of the model's mechanism, and a model deliberately starved of a century of legal and regulatory development is not the same thing as a model that has been tried and found wanting.

What this actually recommends doing today

Put together, the previous essay's negative claim and this one's positive claim yield a specific, non-hand-wavy program, not a shrug.

Where a public safety net exists at all, prefer cash to categories — Friedman's negative income tax over a thicket of in-kind programs, on the evidence GiveDirectly and its peers have since supplied.

Remove the regulatory barriers, largely inherited from an era built to protect large insurers rather than small community pools, that make it needlessly difficult to start a modern mutual aid or micro-insurance association today — a friendly society for gig workers, a sickness fund for a trade guild, a benefit pool for a congregation, all of which face licensing and capital-reserve requirements calibrated for institutions a thousand times their size.

Protect and extend the tax treatment that keeps direct charitable giving cheap rather than funneling more of it through public programs whose overhead and knowledge problem this series has already priced out. And treat platforms that do at internet scale what a lodge secretary did at neighborhood scale — matching a specific, verified hurt to a specific willing giver, the way direct cash-transfer platforms and person-to-person fundraising now can — not as a charming supplement to the real safety net, but as the modern instance of the exact mechanism that worked before it was crowded out.

None of this requires believing hurt isn't real, or that nothing should be done. It requires believing that the question "what should be done" has an answer worth getting right, and that getting it right means putting the resources, the knowledge, and the choice back as close as possible to the person actually hurting — which a check with no strings on it does, and a bureaucracy built to treat every hurt as the same hurt structurally cannot.