Chapter 4 — The Insurance Promise, from Private Misfortune to a Collective Claim
Social insurance did not prevail because societies finally discovered the right formula. It changed the status of certain losses. An accident, illness or loss of income could still strike a particular person without falling entirely on that person and their family. This shift required contributions, categories, procedures and an institution able to recognise the loss. Above all, it turned a request for assistance into a right to claim [1, 2].
Jules is twenty-four in 1897. He has worked since the age of fourteen in the workshops of a shipyard in Nantes. He knows how to file, drill and fit metal parts. He also knows the belts that slip, the machines restarted too quickly and the tasks performed because the foreman is waiting.
His wage supports his wife and their young child. Another child is due in the spring. One November morning, a transmission jams and production slows. Jules approaches with another worker. His sleeve is caught in the mechanism. A few seconds are enough to break his arm.
The injury immediately brings another ordeal. Who was at fault? Should the employer have made the machine safer? Should the foreman have stopped the work? Should Jules have refused to intervene? Everyone knows the habits of the workshop, but those habits are difficult to turn into evidence before a judge.
The employer speaks of carelessness. The family speaks of a task that Jules could hardly refuse. While responsibility is debated, the wage disappears. His wife borrows money, postpones the rent and asks for relief. The accident arose within the organisation of work. Its cost still falls on the household as a private misfortune.
The French law of 9 April 1898 on workplace accidents changed this situation without removing every conflict. It allowed an injured employee to receive compensation without having to prove a specific fault by the employer. The accident became an occupational risk, meaning a loss connected to productive activity and financed through rules prepared before it occurred [1, 3].
This change involved more than paying a sum of money. It determined whom the worker should approach, which facts had to be established, how long the worker might wait and which resources would finance payment. The injured person no longer depended only on a judge identifying a wrongdoer or a benefactor agreeing to help. The worker could assert a claim, a recognised right to require a payment or service under an established rule.1
Insurance acquired political force through this change. It no longer treated the accident only as an encounter between a careless individual and a dangerous machine. It connected the injury to an activity that regularly produced harm and therefore had to bear part of its cost. This socialisation protected workers. It also gave institutions the power to define recognised accidents, covered people, benefit levels and required evidence [2, 4].
Fault, Relief, Rights
Fault remains an essential response to harm. A person who injures another through negligence should be held responsible. The principle becomes harder to apply when loss results from an entire organisation. In a workshop, machinery, pace, fatigue, incomplete training and fear of losing one’s job may all contribute to the accident. No single action then captures its full cause.
Fate offers another interpretation. Fire, illness or early death can be received as misfortunes of the world. This view may inspire compassion, but it does not identify who must pay.
Charity responds by providing money, food, shelter or care. It may save a family. It nevertheless preserves dependence on the person who gives. Assistance remains revocable, and the recipient must often show that they deserve it.
Mutual aid and benefit societies began to change this relationship. Members contributed before knowing which of them would become ill, injured or unable to earn. The recipient was no longer only the object of generosity. They belonged to the group that had prepared the assistance. These organisations could be disciplinary and sometimes excluded people who did not meet their standards. Even so, they turned a favour into a reciprocal obligation [4–6].
The language of risk adds another operation. Accidents must be brought together, their frequency estimated and resources collected before they occur. Attention moves from the singular case to a series.
This shift does not remove morality. Risk also identifies people who supposedly should have anticipated, avoided or protected themselves. A person is judged not only on what they did, but also on their presumed ability to foresee and manage the future [7, 8].
Insurance therefore did not carry society from a moral world into a purely technical one. It reorganised responsibility. It distinguished accident from fraud, covered loss from negligence and reasonable prevention from impossible obligation. Contemporary models continue this history when they turn an address, behaviour or past event into evidence of prudence. Their novelty lies less in judgement itself than in the volume of traces that can now give judgement an impersonal appearance.
Bringing Accidents into a Series
An accident remains a singular experience for the person who suffers it. Insurance places it beside other accidents to estimate a frequency and a cost. During the nineteenth century, expanding statistics on births, deaths, crime, illness and injury revealed regularities that individual stories could not show as clearly [9].
Putting events into a series opened rights. When accidents cluster in particular occupations, places or groups, it becomes harder to attribute them only to the carelessness of their victims. Statistics can show that working and living conditions produce recurring losses. They also give administrations and insurers the means to classify situations, set benefit schedules and process many cases under a common rule [10, 11].
The categories are not labels attached to a world already arranged. Someone must define what counts as an accident, select an observation period, identify a population and decide which cases can be compared. The resulting frequency depends on the offices that receive declarations, the forms people complete and the rules that make two events comparable. Risk is therefore also a way of processing situations. It gathers cases and relates them to a population [12].
This construction does not make loss imaginary. Jules’s broken arm remains real. It requires us to separate two questions. The first concerns measurement. How many accidents occur and how much do they cost? The second concerns use. Which activity should finance the loss, who may claim and what share will remain with the household? Statistics can inform these choices. They cannot settle them alone.
An average may protect and discipline at the same time. It makes collective exposure visible, then sometimes becomes a norm that every file is expected to fit. A person who falls awkwardly within the category may have to provide more evidence. Quantification removes part of the decision from personal judgement, but it can move arbitrariness into the definition of categories, thresholds and exceptions [10, 11].
The Workplace Accident Changes Status
The workplace accident was one of the first major sites of this transformation. The question was no longer only who had acted wrongly. It became what an industrial society owed to those injured by its industry.
Germany introduced compulsory workplace accident insurance in 1884. The United Kingdom adopted legislation in 1897 and France followed in 1898. Ontario created its system in 1914. In the United States, reform proceeded state by state during the early twentieth century [13–16].
These systems were not identical. Some relied on private insurers, others on funds or public administrations. Courts, employers and worker representatives were given different roles. All responded to the same difficulty. Requiring an employee to reconstruct a precise fault left many injuries uncompensated. The cost of industrial activity then fell on the families least able to absorb it [2, 16].
Recognising occupational risk made this cost more predictable for employers. It also pacified part of the conflict by replacing some litigation with schedules and procedures. That pacification was not neutral. A struggle over employer responsibility could become a technical debate about contribution rates or degrees of disability. Social insurance protected workers while administering their injuries [1, 4].
The reform also recalled that labour cannot be separated from the person who performs it. Karl Polanyi described labour as a fictitious commodity because it is not produced for sale like an ordinary object. It is carried by human beings whose lives do not stop when employment becomes dangerous [17]. Treating workplace injury as a cost of production makes the company and the wider collective bear part of the harm created by the activity.
This reasoning extends beyond the nineteenth-century factory. Roads produce collisions. Planning decisions concentrate some flood losses. Digital systems can spread the same failure across many organisations. Individuals experience the damage, but its conditions were often created long before the event. The same difficulty returns whenever causation is dispersed and an institution must decide who will bear the uncertainty.
In the spring of 2020, workers’ compensation systems encountered a difficulty that older factory machinery had made less visible. A nurse, firefighter or other essential worker could contract Covid-19 after many possible exposures. It was often impossible to establish exactly when and where infection had occurred.
Requiring precise proof could preserve a right in theory while making it almost impossible to use. Several American states therefore created a presumption of occupational origin for certain groups. A presumption is a legal rule that treats a fact as established unless sufficient contrary evidence is produced.2
The rule did not state that every infection had been contracted at work. It changed the burden of proof. The employee no longer had to reconstruct an invisible transmission. The employer or insurer could still dispute the occupational connection, but had to provide evidence for doing so.
This shift immediately raised another question. How many workers would be covered and what would the cost be for compensation systems? The debates of 2020 returned to the conflict of the late nineteenth century. When individual causation cannot be proved with certainty, should the uncertainty remain with the person who became ill or be carried by the activity that exposed them more heavily? [18, 19]
From Assistance to a Claim
Social insurance changes the position of the recipient. Under charity, the person asks for assistance that depends on discretion. Under a right, the person can invoke a rule against the institution. The patient, injured worker, unemployed person or pensioner becomes entitled to a benefit defined in advance. Solidarity no longer depends only on the emotion inspired by an individual story [20, 21].
This claim gives autonomy a material basis. Income replacement, a pension or payment for care can prevent an accident from leading immediately to loss of housing, dependence on relatives or lasting exclusion from work. Robert Castel used the expression social property for protections that give people without substantial assets some of the security once reserved for property owners [22, 23].
Protection does not connect an isolated individual directly to the state. Mutuals, occupational funds, unions, employers, municipalities and insurers stand between them. European health insurance systems developed partly from occupational funds and associations before public authorities supported, required or consolidated them [24]. In the United States, employment based benefits provided important protection while giving large firms a central role in access to social rights [25].
Socialising a risk therefore does not necessarily mean assigning all protection to a central administration. It means building a chain of obligations strong enough to prevent loss from falling directly on the household. Every intermediary may protect while also creating dependence. The continuing questions are who governs the institution, who is represented and who remains outside its reach.
The payments that finance this chain do not organise the same relationship. A private premium buys a contractual promise and reflects expected losses, expenses, resources for bad years and sometimes profit. A social contribution finances a common scheme and may depend more on income than on individual expected cost. Taxation supports a political obligation without requiring a direct correspondence between what each person pays and receives.
These forms mix in practice. Private insurance depends on public rules. A social scheme uses actuarial calculations. A state guarantee may support reinsurance, and a mutual may operate much like a large company. The distinction still matters. The same payment may purchase a contract, open a right or mark membership in an institution. Reducing all three relationships to the question of individual cost would erase what they organise in common.
Law Protects and Filters
Turning relief into a right does not guarantee that everyone will receive the benefit. A person must enter the correct scheme, report the event, provide the expected documents and have the case recognised under the relevant definition. Administration protects and filters in the same movement. The claimant no longer depends on the compassion of a benefactor, but does depend on a qualification.
Qualification may be simple when the accident is visible and the rule clear. It becomes harder when several causes combine, symptoms emerge gradually or the person does not understand the procedure. A right can then exist without being practically accessible. Learning the process, waiting for a decision and pursuing an appeal require time, money and sometimes specialist help. People with the fewest resources are also most likely to abandon the process before receiving what they are owed [20, 23].
The insurance market itself depends on legal rules. Bernard Harcourt reminds us that a market never exists before the institutions that define contracts, participants and sanctions [26]. In insurance, law determines who may sell a policy, which information may be used, how a contract may be cancelled and which appeals are available. Prudential rules also require insurers to retain enough resources to meet future commitments.
The state is not external to insurance history. Insurers helped build instruments of credit, information and finance that public authorities also used. Hannah Farber shows how marine insurers contributed to the financial architecture of the early United States [27]. Law makes contracts enforceable. Insurance techniques, in turn, allow companies and governments to undertake commitments in the face of uncertainty.
This organisation retains an ambiguity. Insurance moves away from fault by recognising that loss may arise from collective exposure. It can bring fault back through prevention requirements. Making a machine safer, maintaining a building or financing early care can genuinely reduce danger. Prevention becomes punitive when a person must prove, alone, a degree of prudence that their income, work or housing does not allow. An obligation that is reasonable for a large company may be impossible for a household or insecure worker [7, 8].
The passage from private misfortune to a collective claim is therefore never completed by legislation or by signing a contract. The claim must still be recognised in an individual file. The policyholder contributed before the loss. After the accident, they depend on the institution that verifies coverage, assesses damage and decides when payment will be made. The next chapter examines this unusual relationship in which the claimant is legally a creditor but rarely controls the procedure that gives the claim practical value.
References
In this chapter, a claim means a recognised entitlement to require a payment or service from a person or institution. In insurance, the word also commonly refers to the request submitted after a loss.↩︎
In law, a presumption allows one fact to be treated as established on the basis of another known fact. Some presumptions can be rebutted when the opposing party provides sufficient evidence.↩︎