Why a church charity knows things a federal program never will

Picture a woman in a mid-sized American city. This is a hypothetical, and I label it as one. She has two children, a car that fails every other month, a job that cuts her hours without warning, and a landlord who has just sent a notice. She needs help. Where does she go?

She can go to a government office, where a caseworker will ask her to fill out a form, document her income, wait for review, and learn whether she meets the eligibility rules. Or she can go to the church down the street, where someone will ask her name, and then ask her what is actually going wrong. The two answers are different, and so are the results. The difference is not about kindness. It is about knowledge.

The claim

Let me state my thesis clearly, then argue for it. Private charity, and church charity in particular, does certain things better than any federal program can, because it knows the individual, it asks something of him, and it lasts.

I am not saying government has no role. I will end with a division of labor that I think conservatives can defend and that gives government its due. But I am arguing that when we judge help by whether it actually changes a life, the small and personal institution often has advantages that the large and impersonal one cannot copy, however well it is run.

Three premises carry the argument.

1. Effective help depends on knowledge of particular circumstances that a central program cannot hold.
2. Effective help usually involves a relationship and an expectation, which a program with rules and entitlements finds hard to provide.
3. Effective help needs to last as long as the need does, and private institutions are more durable than grant cycles.

Hayek and the problem of knowledge

In September 1945 the economist Friedrich Hayek published an essay in the American Economic Review called "The Use of Knowledge in Society." It is one of the most important papers in twentieth-century economics, and it is not primarily about charity at all. It is about the central problem of any economy, which Hayek said is not how to allocate known resources among known ends, but how to use knowledge that is dispersed among millions of people and not given to anyone in its totality.

Hayek stressed one kind of knowledge in particular: what he called the knowledge of the particular circumstances of time and place. The shopkeeper who knows that a certain item is in short supply on his street, the dispatcher who knows that a truck is idle, the neighbor who knows that a family on the block lost a job last month. This knowledge is local, perishable, often unspoken, and impossible to gather into a statistical report. A central authority cannot use it, because it cannot know it.

Apply that to relief. A federal program works from categories: income below a threshold, a family of a certain size, a particular qualifying condition. Those categories are the only knowledge that can travel up to a central office and down again in the form of a rule. But the woman in my hypothetical does not fit a category. Her problem is a car, a schedule and a landlord. The thing that would help her may be two hundred dollars for a repair, a ride to work for three weeks, and a conversation with her landlord, none of which is on any form.

A neighbor or a church member can learn this, because they ask, and they live nearby, and they can check back. A federal program cannot, because the information that would make the difference is the very information that its structure screens out. For a plain-language introduction to Hayek's wider argument, see The Austrian school in plain words, from Carl Menger to Friedrich Hayek.

What Olasky found in the nineteenth century

In 1992 Marvin Olasky published The Tragedy of American Compassion, a history of American poor relief from the colonial era to the modern welfare state. His central observation was that the nineteenth-century charity workers he studied, in churches and in organizations such as the charity societies of the period, practiced a style of help that the twentieth-century welfare state largely abandoned. They distinguished between people who needed temporary material help and people whose troubles were deeper, and they tailored their help accordingly. They asked people to work in exchange for assistance where they could. They tried to involve the recipient's family, neighbors and employer. And they saw the spiritual and moral dimension of a person's difficulty as part of the problem, not as an intrusion.

Olasky's account has critics, and you should read them. Some historians argue that he idealizes the older charities, which had their own harshness and their own failures, and that private relief was never adequate to the largest economic shocks. I will come back to that. But his core point survives the criticisms: help that is personal, discerning and expects something of the recipient is a different thing from a check, and in many cases a better thing.

For a historical example of a charity that fits the pattern, see William and Catherine Booth, the Salvation Army and soup, soap and salvation, which tells the story of an organization that fed the hungry and told them the truth in the same breath.

Brooks and the data on giving

In 2006 Arthur Brooks, then a professor at Syracuse University, published Who Really Cares, a study of charitable giving in the United States. Its most discussed finding was that religious Americans give more, and give more often, than secular Americans, and that the pattern extends to non-religious causes: people who belong to a congregation are more likely to give to secular charities and to volunteer for them than people who do not. Brooks also reported that, in his analysis of survey data, households headed by conservatives gave more than households headed by liberals, even after controlling for income.

I want to be careful here. Brooks's findings come from survey data and they are an argument, not a proof. Later researchers have disputed some of the details, including whether the gap between conservative and liberal households survives corrections to the data, and Brooks himself traced much of the difference to religion rather than politics. Measuring generosity is hard. But the broad point, that religious practice is strongly associated with giving and volunteering, has been found in a good deal of other work, and you can check the book and its sources yourself. The relevance to my argument is this: churches and congregations are not merely one more charity. They are a mechanism for producing the giving and the volunteering that charity runs on.

Asking something of the person

The second premise is the one that makes people uncomfortable, so let me say it directly. Good charity often asks something of the person it helps.

That is not cruelty. It is respect. A federal program that determines eligibility by income and gives the benefit without any expectation treats the recipient as a case to be processed. A church charity that says, in effect, "we will help you, and here is what we hope you will do, and we will check in," treats him as a person who is capable of change and responsible for his own life. It offers help and a relationship. And the relationship has a hold on the person that a form does not.

Consider what that can mean in practice. A congregation that helps a man who is out of work may also connect him with its members who own businesses. It may ask him to attend a class, or to commit to a recovery program, or simply to come back next week and say how it went. The expectation is part of the care. A government program, by contrast, is bound by rules that treat conditions as intrusions. Constitutional and statutory constraints, not unlike the ones that govern the conduct of any public agency, limit what an official may ask of a citizen about his habits, his faith, or his family. Those constraints are reasonable for a government. They are also exactly why government cannot do certain kinds of helping.

This is also where one of the quiet scandals of the modern system lies. Programs that phase out benefits as income rises can penalize work and marriage. the marriage penalty buried in welfare programs, with the math shown works through the arithmetic. A church that is trying to help a family stand on its own feet does not have a phase-out schedule.

Outlasting the grant cycle

The third premise concerns time. Federal programs are funded in cycles. A grant is awarded for a period, a program is evaluated, and the funding may be renewed, shifted or cut. Administrators change, priorities change, and rules change. A nonprofit that depends on one grant is a hostage to the next appropriations bill.

A congregation does not live that way. Many churches in American towns have been in the same place for a century or more. They have an institutional memory and a stable base of members. A family that was helped ten years ago may now be among the volunteers. The relationship does not end when the paperwork period expires, because the relationship was never created by the paperwork.

I am aware that this is a generalization, and that some churches are small, short-lived or badly run. Charities fail. What I am saying is that the form of the institution gives it a staying power that a program defined by a funding line does not have.

The steelman: only government works at scale

Now the strongest objection, which I take seriously. It goes like this.

"All of this is lovely, but it describes help at the scale of a neighborhood. Poverty in a modern economy is a problem of millions. In the Great Depression, private charity and local relief were overwhelmed, and it was the federal government that put food on tables and built a safety net. Retirement, disability and unemployment are risks that strike unpredictably and at scale, and pooling those risks requires a large institution. Churches are unevenly distributed, and a poor community often has the poorest churches. Charity depends on the generosity of the nearby, while need does not respect geography."

That is a good argument, and a conservative who ignores it is not being serious. Let me concede what is true in it.

• Scale matters for certain risks. Insurance against catastrophic and widespread events, such as a national recession, a disaster, or the broad risks of old age, is something a pooled and large institution can do and a congregation cannot.
• Private charity has limits in a deep crisis. The Depression showed that. Local and private relief was not designed for unemployment on that scale, and it gave out.
• Uneven distribution is real. A poor neighborhood may have little capacity to help itself.

I accept all three. But the argument has less force than it seems, for reasons I will give in order.

1. Scale is a reason for government to do some things. It is not a reason for government to do everything. The argument defends a social insurance function. It does not show that a program that supports the ordinary working poor, with the personal questions and local circumstances I described, should be run from a central office.
2. Before the welfare state, Americans built an astonishing network of mutual aid. Fraternal lodges, benevolent societies, churches and unions provided insurance and support to millions of people. Before the welfare state, Americans insured each other through lodges describes that world. The crowding out of those institutions by government programs is part of the story, and not the happiest part.
3. The relevant question is the margin. Nobody says government should do nothing. The question is whether, at the margin, a further dollar or a further program is better spent through a government channel or through a local and personal one, for a given kind of need.
4. Government can fund and still delegate. In 1996 Congress included in its welfare reform law a provision, often called Charitable Choice, that made it easier for faith-based organizations to take part in government-funded programs without giving up their religious character. Whatever its limits, the provision reflected the view that government can use the knowledge and relationships of private institutions rather than replace them.

A division of labor conservatives can defend

So here is the arrangement I would propose, and I think it follows from the principle that subsidiarity describes. Subsidiarity, the 1931 Catholic principle every federalist should know sets out the idea: a higher body should not do what a lower one can do, and should help the lower body when it falls short.

• Government handles what is large, predictable in kind and best insured by pooling. Defense, the courts, protection against catastrophic and economy-wide risks, and a basic backstop for those who cannot be helped in any other way.
• Civil society handles what is personal. Counseling, recovery, mentoring, job connections, help in the first weeks of a crisis, and the long work of restoring a person to a stable life.
• Government stays out of the way of the second function and, where it funds anything, funds the local and personal institutions without smothering them. That means light conditions, fair access for faith-based providers, and a willingness to let the local institution do what it does.
• Citizens fill the space. If we say that private charity does the work better, then we owe it to the argument to give. How to Give Well, a Family Playbook for Charity That Actually Helps offers a practical way to start.

This division does not abolish the safety net. It asks that the net be made of different materials in different places. For the larger principle that conservatives are not against government but against rule by bureaucracy, see Conservatives Are Not Against Government, We Are Against Rule by Bureau.

What Tocqueville saw

Alexis de Tocqueville traveled the United States in 1831 and 1832 and published Democracy in America in two volumes, in 1835 and 1840. One thing that astonished him was the American habit of forming associations. Americans of all ages and conditions, he wrote, were constantly forming them, for commerce, for religion, for morals, for serious and trivial purposes. Where a Frenchman would look to the government, Tocqueville said, in the United States one would find an association.

That was the American answer to the problem of how free people take care of one another without a master. It still is. The test of our principle is whether we are willing to take up the work, because a society that talks about private charity and leaves the work to the state is like a man who praises exercise from the sofa.

The case in brief

1. Effective help depends on local knowledge, as Hayek argued in 1945, and a central program works from categories that cannot capture the particular circumstances of a person's trouble.
2. Church charity can ask something of the person it helps and offer a relationship, and Olasky's history of nineteenth-century charity and Brooks's study of giving suggest this combination has real power.
3. Congregations outlast grant cycles, because the institution does not depend on a funding line.
4. The strongest objection, that only government works at scale, is right about catastrophic and pooled risks, but it does not show that personal and local help should be run from a central office.
5. A defensible division of labor lets government insure against the large and predictable, and leaves the personal work to churches and other private institutions that government should fund lightly and not displace.

What you can do: pick one church, charity or mutual-aid group in your own community that helps people in need, ask what it needs most, and give it money, time or both this month.