Skip to content

Assurance Contract

Conditional-commitment protocol — instantiates Public Goods Provision

Collects conditional pledges that bind only once enough others also commit, so no one has to pay first for a good that needs a crowd.

An Assurance Contract breaks the "I'll pay if you pay" standoff that keeps lumpy shared goods unfunded. Each participant makes a conditional pledge; the pledges are held in escrow and become binding only when total commitments cross a preset threshold before a deadline. If the threshold isn't reached, everyone is refunded and no one is out anything. Its defining move is that the threshold itself is the free-rider defense: because your money is charged only if the good actually gets funded, you can never be the lone sucker who paid for a bridge that was never built. That turns "wait and hope someone else pays" into "pledge now, risk-free."

Example

Forty households share a private gravel road that needs repaving — roughly $80k of work. Passing the hat has failed twice: each family reasons that its own $2,000 is wasted unless everyone else pays too, so everyone waits. A neighbor sets up an assurance contract instead. Every household pledges its share into escrow, with one condition written in: no card is charged unless pledges reach the full $80k by the end of the month; short of that, all pledges are voided. Now the calculation flips — pledging costs nothing if the effort fails, and the only way to get the road is to commit. Holdouts lose their reason to wait, the threshold is cleared in three weeks, and the escrow releases to the contractor.

How it works

What sets this apart from an ordinary collection drive is the binding condition, not the ask:

  • Conditional escrow. Pledges are collected but not charged; they sit in escrow, reversible until the threshold decides their fate.
  • A hard threshold and deadline. A specific funding target and cut-off date; cross it and all pledges fire together, miss it and all reverse together. Nobody moves alone.
  • All-or-nothing settlement. There is no partial provision — the good is funded in full or not at all, which is what makes the pledge safe.
  • Optional dominant variant. The organizer can promise pledgers a small bonus if the threshold fails, making it individually rational to pledge regardless of what others do.[1]

Tuning parameters

  • Threshold level — set it to the good's true full cost and it protects contributors; set it low to "make the number," and you fund a token version that underdelivers.
  • Refund-vs-penalty terms — plain refund on failure is simplest; a failure bonus (the dominant variant) recruits fence-sitters but costs the organizer if it flops.
  • Pledge deadline — a tight window forces decisions and momentum; too tight and legitimate contributors can't organize in time.
  • Pledge form — fixed equal shares feel fair and simple; pay-what-you-can raises more from those who value it most but complicates the threshold math.
  • Pledge visibility — public pledges create social proof and pressure toward the threshold; private pledges protect people who don't want their contribution known.

When it helps, and when it misleads

It shines for one-time launches of a lumpy, well-priced good where the real obstacle is coordination, not generosity — the threshold manufactures the mutual assurance that voluntary asks lack. Its blind spot is afterward: an assurance contract funds a creation event and says nothing about upkeep, so a good launched this way can be built and then left to rot. It also needs a genuinely crisp total cost; where the price is fuzzy or open-ended, there is no clean threshold to set. The classic misuse is lowballing the threshold to guarantee a "success" — which merely converts an honest under-provision into a funded-but-inadequate one. The discipline is to set the threshold at the good's real provision cost and to pair the launch with a standing maintenance model from day one.

How it implements the components

  • assurance_threshold — the mechanism is the threshold: the commit-only-if-enough-others-do condition that gates every pledge.
  • contribution_rule — defines each participant's conditional share and the terms under which it becomes owed.
  • funding_pool — the escrow aggregates pledges into a pool that releases as usable capacity only at threshold.
  • free_rider_response — all-or-nothing settlement removes the incentive to hold out, since waiting no longer offloads cost onto others.

It does not implement the upkeep components (maintenance_obligation, reserve_or_endowment) — those belong to Maintenance Endowment or Reserve — nor the standing governance component (provision_responsibility), which is Cooperative Ownership's.

Notes

An assurance contract launches; it does not sustain. Left alone it produces exactly one funded event, so treat the threshold as the start of provision and attach a recurring contribution or reserve mechanism before the escrow releases — otherwise the maintenance obligation lands on whoever happens to care most.

References

[1] A dominant assurance contract adds a payoff to pledgers if the threshold is missed, so pledging becomes the individually rational choice no matter what others do — Alexander Tabarrok's refinement of the classic assurance contract for public goods. ↩