Every enduring institution has had to answer the same question: what does a member get to keep when they leave? The answer is rarely incidental. It shapes who joins, how power is exercised while they stay, and whether the institution can be trusted with anything worth guarding.

Consider the range of historical answers. A medieval guild admitted a member for life and offered almost nothing on departure; leaving meant forfeiting the trade itself, which is precisely why guilds could enforce their standards. A worker cooperative, in the tradition Elinor Ostrom spent her career documenting, ties membership to labor and returns a member's patronage stake when they go – the exit is real, but it is a settlement of accounts, not a claim on the whole. The joint-stock company drew the sharpest line of all: a shareholder may sell at any moment, to anyone, at whatever price the market will bear, and the firm neither notices nor consents. Liquidity replaced permission. That single move – making the share severable – is most of what made the modern corporation financeable.

Each of these designs pairs an exit rule with a governance rule, and the pairing is never accidental. Where you cannot leave with anything, voice tends to be loud and votes tend to be cheap, because dissatisfaction has nowhere else to go. Where you can leave with everything, voice is often thin, because the disciplined threat of departure does the work that argument would otherwise have to. Albert Hirschman named these the two responses available to any member of any organization – voice and exit – and observed that the health of an institution depends on the balance between them. Most organizations get to choose that balance. What they cannot usually do is make the two quantities the same.

This is the design question worth dwelling on, because digital organizations have mostly answered it badly. The first wave of on-chain organizations imported one-token-one-vote from nowhere in particular and discovered its failure mode within a season: a member's voting power and their financial claim came apart. You could accumulate votes without committing capital that anyone else could redeem, and you could hold a redeemable claim while never showing up to govern. Power and stake, decoupled, produce exactly the pathologies you would predict – governance capture by the well-capitalized, apathy from everyone else, and treasuries that belong to the organization in name but to no member in any way they can act on.

A cleaner answer is to refuse the decoupling outright: make the share a member can leave with equal to the weight their vote carries. Not correlated, not roughly proportional – identical. Under such a rule you cannot vote with more force than you could walk out the door with, and you cannot hold a larger claim on the treasury than your voice reflects. Voice and exit stop being two levers to be balanced and become one quantity seen from two sides.

Caper is built on that identity. A member's share of the common treasury, at the moment they choose to leave, is the same figure that determines how much their vote counts while they stay. The weight is a function of what they actually hold and the participation stake they have committed, measured against the whole; the treasury share on exit is that same weight, applied to the same treasury. When a member exits, they take that fraction of the common funds and their claim is retired – the exit is genuine severance, in the joint-stock sense, not a promise to be settled later. What is new is not the liquidity; markets have offered that for centuries. What is new is that the liquid, severable claim and the governance weight are guaranteed by construction to be the same number.

The consequences run in both directions, and both are the point. A large holder cannot out-vote a small one beyond the share they could actually redeem, which caps the return on governance capture at the return on simply cashing out – and a holder who would rather cash out than steer is welcome to. A disengaged member is not trapped subsidizing a treasury they have no voice over; the door is open, and it is open at their real weight, not at some haircut the majority sets on the way out. The exit right, priced honestly, disciplines the people who stay. It is the credible threat that makes voice worth listening to, because everyone in the room knows exactly what walking away is worth.

There is a further reason this matters now, beyond the governance of organizations that already exist. If the severable share and the vote can be made one number for a group, they can be made one number for a person. An individual who incorporates themselves – issuing a claim on their own future, backed by a treasury, governed by the people who chose to back them – needs precisely this guarantee to be trustworthy. The personal-token experiments of the last decade failed for want of it: they offered a claim with no governance and no honest exit, and when trust wavered there was nothing underneath. A treasury and a real exit right are what make self-incorporation credible rather than speculative. The exit right is not only how a member leaves an organization well. It is the condition under which a person can become one.