Chapter 2 — What Makes the Pool Hold Together
An institution can calculate with great precision what each member costs the group and still lose sight of why those with lower exposure would agree to remain with everyone else. Pooling does not depend on uncertainty alone. It also rests on rules that make contributions intelligible, prevent individual departures from dismantling common protection and distribute obligations among policyholders, companies and public authorities.
For years, the people living on a plateau and those living in a valley had been covered by the same mutual insurer. Homes on the plateau rarely flooded. In the valley, water reached basements and ground floors more often. Contributions reflected the difference, but only in part.
No one described the arrangement as a transfer. Each household received its own bill, and some later reported a claim. Then a new map made it possible to estimate exposure almost address by address. A competing insurer offered a substantial reduction to households on the plateau.
Several members asked their mutual to match the price. They did not say that they wanted to abandon the valley. They simply wanted to pay an amount that corresponded more closely to the risk of their own home.
The board divided. Reducing contributions on the plateau could require a steep increase in the valley. Refusing the reduction might cause the least exposed members to leave, forcing those who remained to carry an even larger share of the losses.
One board member asked why the people on the plateau should continue to contribute more. They might face a different risk tomorrow, another member replied. The two territories shared roads, public services and the same river, said a second. A third asked what would happen to homes in the valley if insurance withdrew. A final member argued that none of these reasons entitled the mutual to preserve a transfer that no one had clearly discussed.
The map had made the difference visible. It had not chosen the common rule.
The previous chapter examined who owns the pool and who has the power to allocate its resources. That question remains important, but it is not enough. Even in a mutual owned by its members, each person may prefer an immediate reduction to a solidarity whose benefits remain uncertain. An institution does not endure simply because its members own it. It must give them reasons to keep financing a promise that may mainly benefit other people.
Uncertainty provides one reason. People contribute because they do not yet know whether they will experience a fire, illness, accident or flood. This solidarity through chance makes pooling easier. It becomes less obvious when data distinguish exposures more closely and some policyholders know that they are likely to remain less costly than others [1, 2].
Altruism provides another answer. Less exposed members may support more vulnerable people because they care about them. Generosity exists, especially after disasters. It is not enough to finance an institution year after year. Policyholders do not know most of the people whose claims their contributions will pay. They must be able to support common protection without feeling personal concern for every beneficiary.
Anonymous blood and organ donation show that an institution can organise a relationship among strangers. Donors do not necessarily choose the recipients. Rules collect, check and distribute the resource [3, 4]. Insurance is not a gift, because a contribution creates a possible right. Yet the arrangements share an important property. The person who suffers the loss does not have to persuade each contributor separately that they deserve help.
Sociologists speak of generalised exchange when assistance does not return directly from the person who received it. One person contributes for another, who may later contribute for a third. Reciprocity passes through the group rather than taking place face to face between two individuals [5]. The pool turns this chain into lasting rights and obligations. It holds when members understand their contribution as financing common protection rather than as money taken from them for the benefit of strangers.
Solidarity without Heroism
Cooperation means producing together a form of security that no participant could obtain alone [6, 7]. Few households can keep enough savings to rebuild a home, replace income for a long period or pay for very costly care. A pool gives them access to a collective capacity to pay.
This interdependence does not remove conflict. Someone who has never made a claim may regret the contribution after the fact. Someone who knows that their exposure is low may look for a cheaper contract. A claimant wants prompt payment, while the institution must retain enough resources for future demands. A common interest does not make every position identical.
Cooperation does not require all members to bear the same cost. It requires differences to be governed by a rule that could still be defended if each person’s position changed. Someone may live on the plateau today, be in good health or hold a stable job. Tomorrow they may move, become ill or lose protection tied to employment. A rule appears less just when it convinces only those who already know that it places them on the favourable side.
The legal philosopher Ronald Dworkin asked us to imagine the insurance people would choose before knowing their health, disability or future ability to earn an income. This hypothetical insurance market does not describe an actual market. It asks what protection people might reasonably choose before knowing whether they would belong to a costly group [8, 9].
The thought experiment does not automatically turn protection into a premium. Nor does it replace rights. A society may have to guarantee care, a minimum income or continuity of housing even when it cannot prove that every citizen would have bought such cover. The experiment nevertheless rules out an overly simple objection. Durable solidarity does not require every contributor to expect to receive more than they pay. It can rest on a rule that each person would have had reason to accept before knowing their own trajectory.
Cooperation also needs recognised positions. A person becomes a policyholder, member, contributor, claimant or creditor because an institution attaches rights and duties to that status [10, 11]. They no longer ask strangers for relief. They demand performance of a promise financed in advance.
This common status can remain abstract. Members need not share an occupation, religion or neighbourhood. They must still be able to understand the principles of sharing. Which differences alter the contribution? Which remain pooled? How are very large risks financed? What limits an abrupt reclassification?
Understanding these principles does not require public disclosure of diagnoses, incomes or addresses. Transparency should concern the rules and their collective effects. An institution can show what share of contributions pays claims, operating expenses and reinsurance, which groups face the largest increases, how many contracts are not renewed and how much is invested in prevention. It can make sharing visible without presenting every beneficiary as personally indebted to the rest of the pool. Participation then depends less on an abstract appeal to commitment than on the ability to understand a decision and exercise real influence over it [12].
In 1882, the 102 farms of Canillo, in Andorra, created a mutual fire insurer called La Crema. The roads through this mountainous region made outside firefighting assistance difficult. The organisation had to coordinate fire protection and compensate members whose buildings were destroyed [13].
The members met each year in an assembly. Each declared the value of the home, barn, stable and other buildings they wanted to protect. These amounts were recorded in several registers. When a fire occurred, compensation could not exceed the declared value. The other members financed the payment after the loss, each contributing in proportion to the value they had declared for their own property.
The rule was easy to understand. Declaring a higher value increased the possible compensation, but it also increased the amount owed when a neighbour suffered a fire. Elected commissioners assessed the damage, monitored building maintenance and managed firefighting equipment. The common council then approved the payment.
La Crema was not free of conflict or strategic calculation. Members could be tempted to overstate or understate the value of their property. Its longevity nevertheless shows what a visible rule can make possible when it is connected to an assembly, local knowledge of the buildings and reciprocal obligations. The contribution did not arrive as a price produced at a distance. Members could relate it to another member’s loss and to a procedure known by the group [13].
A small collective such as La Crema makes relationships easier to see. It is not a model that can simply be transferred to millions of policyholders. Proximity helps members know the buildings and monitor conduct. It can also intensify social surveillance, exclusion and pressure on those who depart from the group’s norms. A larger pool spreads losses more widely and reduces some of the transactions needed to share them. It also makes it harder to connect contributions to beneficiaries [14–16].
Insurance must therefore connect several scales. Diagnosis, some prevention work and part of the deliberation can remain close to the people concerned. Capital, reinsurance and protection against catastrophe often require a much larger population. Subsidiarity refers to this search for the appropriate level of decision. It does not mean abandoning each group to its own resources. It places a decision where it can be understood and challenged, then distributes losses at a level broad enough to carry them.
Cooperation Must Be Protected
A willingness to cooperate does not survive on its own. It depends on what each person expects others to do and on whether the institution can prevent some actors from benefiting from the pool during favourable years, then leaving as soon as their contribution rises [17].
A member pays today because they expect the promise still to be financed when they need it. This confidence rests on observable practices. Are recognised claims paid without unjustified delay? Do managers explain how resources are used? Is proven fraud addressed without turning every claim into suspicion? Are rules changed with enough notice and a clear explanation?
Control can protect the common fund. Verifying a declaration, requesting evidence or sanctioning demonstrated fraud prevents a person from receiving money to which they are not entitled. Sanction can also reassure members who are willing to contribute because they believe that the same obligations apply to others [18].
This discipline loses its justification when it falls only on the least powerful member. The policyholder is responsible for accurate declarations and for prevention measures genuinely within reach. The insurer is responsible for continuity of cover, solvency and fair claims handling. The owner, municipality, developer and network operator must answer for decisions that create or maintain exposure. An institution destroys the trust it claims to protect when it monitors small individual actions precisely while allowing powerful actors to shift their costs into the pool.
Competition makes this protection more difficult. An insurer can attract less costly people with a lower premium. The institution left with more high risks must then increase its prices, which encourages further departures. This mechanism is called adverse selection when those who remain, or buy more cover, are on average more exposed than those who leave [19, 20].
Compulsory insurance can limit this flight by keeping a broad population in the same system. It is not enough when companies still gain from selecting the least costly members. Transfers among insurers can then direct resources towards organisations that cover more high-risk people. This mechanism, usually called risk adjustment or risk equalisation, aims to make firms compete over service, expenses and the quality of prevention rather than over their ability to avoid costly members [19, 21, 22]. European social health insurance systems have developed several forms of solidarity and equalisation to preserve a common arrangement despite the presence of several funds or insurers [23].
No equalisation mechanism removes every difference. A model can underestimate some needs and leave insurers with an incentive to discourage people for whom compensation is inadequate. It can also become so complex that only a small group of specialists understands its effects. The principle remains important. Freedom to choose a provider does not imply freedom to dismantle all solidarity. Several providers can operate within common rules that remove at least part of the gain from selection.
Stability also requires time. Members may contribute for years without receiving a payment. Reserves were built by contributors who have since left the group. Prevention work will benefit future residents and policyholders. The pool therefore connects several generations of members. A decision confined to the current accounting year may distribute accumulated resources too quickly or reject an investment whose benefits will appear later.
Price Changes the Relationship
A premium, discount or excess does more than alter the policyholder’s budget. It also gives meaning to their relationship with the group. Incentives send messages as well as changing economic choices [24].
A discount after preventive work may recognise a genuine contribution to common security. It may encourage the installation of a backflow valve, reinforcement of a roof or safer driving. It may also suggest that each person alone owns the benefits of their prudence and that any remaining vulnerability should be borne as a personal fault.
Conversely, a contribution that never recognises effort may discourage people who have paid for useful work. Cooperation does not require institutions to ignore conduct. It requires them to distinguish an accessible action that genuinely reduces risk from a social constraint merely recorded as individual behaviour. Effort can be recognised without treating the observed outcome as a complete measure of merit.
Personalisation also changes how policyholders imagine the other members. Someone whom the model labels a good risk may feel wronged whenever they pay more than their expected cost. Someone who receives an increase may be presented as the beneficiary of a favour granted by everyone else. Relations of solidarity then become hidden subsidies that ought to be removed [25, 26].
This reading forgets that insurance exists precisely to separate the present contribution from the future loss. People pay before knowing who will need the resources. The contract creates a right without requiring the claimant to obtain the personal approval of every contributor. It protects the claimant from dependence on charity.
The institution can still reintroduce a test of merit through exclusions, controls and conditions of cover. The policyholder must prove that they belong to the accepted category, declared the risk correctly and followed the expected conduct. Such checks are sometimes necessary. They cease to serve the promise when every vulnerability becomes grounds for suspicion or every difficulty in preventing loss is treated as fault.
A captive population requires stronger rights. When insurance is necessary to drive, borrow, work or obtain care, a formal choice among several contracts is not enough. Continuity of cover, justification of the price, minimum quality of protection and a practical route of appeal become central duties of the institution.
When Leaving Becomes Rational
The board of the plateau and valley mutual could not avoid the conflict simply by asking the model for a more accurate price. Every household on the plateau had an understandable reason to accept the cheaper offer. If they all did so, however, the valley would lose part of the contributions that still made insurance possible there.
An appeal to generosity would not solve the problem. The mutual could make the transfer explicit, limit the speed of increases, reserve a common part of the premium for the largest losses and use another part to signal differences in exposure. It could also finance work in the valley so that solidarity did not merely preserve the same danger indefinitely. Each solution would create new disagreements about the amount, duration and beneficiaries of the effort.
The decisive point was not to let a series of individual departures silently produce the collective rule. Competition can improve service and contain expenses. It weakens pooling when it chiefly rewards the ability to attract the least costly members [21]. The common rule must organise freedom of choice without allowing one person’s exit to remove another person’s access to essential protection.
When such a rule is missing, a society may have many insurers and many contracts while leaving a growing share of loss with households. Protection then depends on employment, income, provider networks, legal status and the ability to absorb an excess. The next chapter examines this fragmentation through the American system.