Costly religious demands
A Socratic walk-through of costly religious demands — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do religious groups that demand the most useless-looking sacrifices tend to last longest?
A group asks its members to give up alcohol, wear clothing that marks them out on every street, keep a diet that makes ordinary hospitality impossible, hand over a tenth of their income, and sit through several hours of assembly a week. None of this feeds anyone or builds anything. On the face of it, a group that asked less should attract more members and prosper.
The historical record runs the other way, and that is the thing to explain. Not "why do people put up with it" — the interesting question is why the demands should be useless. If the point were contribution, the group could ask for work. Why ask for waste?
Reasoning it through
REASONING #Start with what such a group actually produces. Mutual aid when you are ill, care for your children, standing in a community, the sense that you are not alone. Notice their common feature: they are enjoyed by everyone present and paid for by whoever shows up. That is the classic public-good shape, and it has a classic vulnerability — someone who attends when it suits, takes the help, and contributes little does better than the person who gives fully. Enough such members and the goods stop being produced.
So the group has a problem, and it is a problem of information, not of will. It cannot see who is committed. Ask, and everyone says yes, because saying yes is free.
What kind of test would work? It has to be something a genuinely committed person is willing to bear and a free rider is not, and the asymmetry must come from the cost falling differently on the two. A useful cost will not do it: if the group demanded work and paid for it, both types would happily comply, since the deal is good for anyone. But a demand that is pure loss — forgone earnings, forgone food, forgone time, social awkwardness — is worth bearing only if you value what membership brings, which is to say only if you intend to stay and participate. The uselessness is not a flaw in the demand. It is what makes it a test.
This is the costly-signalling account, which Laurence Iannaccone set out for religion in economic terms and Richard Sosis and others developed empirically. Its logic is the one running through the collection's piece on conspicuous signals, where an animal's expensive advertisement is credible precisely because a weaker animal could not afford it. Here the currency is commitment and the audience is the congregation. It also gives an economic edge to the point made in ritual meaning: a rite need not do anything instrumental to be doing something real.
Two checkable consequences follow. A demanding group should have more committed members on average, since the demand screens at entry and keeps screening — the tithe is due every year. And demands must be visible or verifiable, or they can be faked: a private affirmation of belief costs nothing and proves nothing, whereas distinctive dress, dietary rules eaten in company, and sabbath restrictions that visibly cost you a day's trade are all observable, which is not a coincidence.
Now the evidence, which is real and narrower than the story. Sosis studied nineteenth-century American communes and found the religious ones markedly more durable than the secular — roughly four times as likely to survive any given year, on his figures. He and Eric Bressler then coded a set of these communes for the number of costly requirements imposed, and found that requirement count predicted longevity among the religious communes but not the secular ones. That is an oddly specific result, and a suggestive one, since it implies the sacrifice needs a sacred frame to bind rather than merely being expensive.
The analogy
THE ANALOGY #Think of an employer who cannot tell, at interview, who will still be there in five years — and for whom the difference matters enormously, because the training is long. So the job opens with a punishing, badly paid first year that teaches almost nothing. It looks like waste, and it is. But someone planning to leave in eighteen months finds it a bad trade, while someone planning a career finds it a small price. The employer never learns anyone's intentions. He simply arranges matters so that only the people he wants find the offer worth accepting.
The employer designs the filter deliberately, whereas nobody sat down and designed a dietary law — the account holds that practices which happened to screen well left more surviving groups behind them, which is a claim about selection over time, not about anyone's intention.
Clarifying the model
THE MODEL #Two refinements, and then an honest accounting of how firm this is.
The screen is not only at the door. Because most demands recur — weekly attendance, annual tithing, lifelong dress — the test runs continuously, which matters more than entry screening for a group that must stay cohesive over decades.
The theory is also not a theory of belief. Its explanatory weight rests entirely on the observable cost, not on the content of what is professed. Sosis's own reading of the commune result is that unfalsifiable belief is what makes a demand non-negotiable, and that is an addition to the signalling logic rather than something it predicts.
And the support, though real, is contested. The commune samples are small and historical, the coding of "costly requirement" involves researcher judgement, and religious and secular communes differ in many ways besides how much they demand — shared origin, leadership structure, external hostility — any of which could drive the survival gap. Causation could also run backwards: an already cohesive group can afford to impose more. There are competing accounts too, notably Joseph Henrich's credibility-enhancing displays, on which costly acts work by making belief itself convincing to onlookers rather than by filtering members. Treat this as a well-motivated hypothesis with genuine but partial support, not a settled finding.
A picture of it
THE PICTURE #How to readA person occupies one state at a time and moves along the labelled transitions. The fork out of "Testing" is the whole mechanism: the same demand is a small price to someone who values membership and a bad bargain to someone who does not, so the two types sort themselves without the group ever learning anyone's intentions. The self-loop on "Committed" is the part usually missed — because the demands recur, the test is re-run continually rather than passed once at the door.
What became clearer
WHAT CLEARED #The apparent uselessness of the demand is doing the work. A useful demand is worth meeting for anyone, so it separates nobody; a costly, visible, pointless one is worth meeting only for those who expect to draw on membership over the long run, so it sorts contributors from free riders without anyone having to read minds. That predicts what the commune data broadly show, while leaving unsettled why the sacred frame appears necessary, and how much of the survival gap the demands themselves explain.
Where to go next
ONWARD #- Why the same logic predicts that groups facing external hostility should tighten their demands rather than relax them.
- Whether the account extends to secular commitment institutions — gangs, elite regiments, medical training.
Key terms
TERMS #| Term | What it means |
|---|---|
| Free rider | someone who takes a group's collective benefits while contributing less than they cost to produce. |
| Costly signal | an act expensive enough that only those with the underlying quality, here commitment, find it worth performing. |
| Screening | arranging conditions so that people sort themselves by a trait the arranger cannot observe. |
| Credibility-enhancing display | Henrich's term for a costly act that makes an onlooker take the actor's stated belief to be sincere. |
Every term the collection defines is gathered in the glossary.