Insurance, Actuarialism and Thrift
DOI link for Insurance, Actuarialism and Thrift
Insurance, Actuarialism and Thrift book
In all the legal struggles to develop a governable category of insurance, there is not a word about insurance (or for that matter gambling or speculation) being deﬁned by its deployment of actuarial or statistical methods. The legislators and the courts appear rather uninterested in the ways in which the insurance industry renders the future calculable. Yet insurance is taken by many theorists as emblematic of the technology of risk. For Beck (1992) the question of whether an insurer will agree to insure is indicative of whether the issue ‘really’ concerns a technology of risk or a technology of uncertainty. Disagreements between scientists who make claims about low levels of risk and insurers who refuse to insure these risks are held to be symptomatic of the currently problematic status of risk-calculative expertise. In turn, this is taken to demonstrate the arrival of risk society. For governmental analysts likewise, risk technology is usually taken to be the principal deﬁning point of insurance:
Risk is calculable. This is the essential point, whereby insurance is distinct from a bet or a lottery. For an event to be a risk, it must be possible to evaluate its probability. Insurance has a dual basis; the statistical table which establishes the regularity of certain events, and the calculus of probabilities applied to that statistic, which yields an evaluation of that class of events actually occurring … When put in the context of a population, the accident which taken on its own seems both random and unavoidable (given a little prudence) can be treated as predictable and calculable. One can predict that during the next year there will be a certain number of accidents, the only unknown being who will have the accident, who will draw one of existence’s unlucky numbers. (Ewald, 1991, p 202; emphasis in original)1
From this basis, Ewald builds up a framework for analysing insurance. The abstract ‘technology of risk’ is distanced from the specific institutions of insurance (for example marine insurance, life insurance, socialised unemployment insurance) by the fact that ‘[i]nsurance institutions are not the application of a technology of risk; they are always just one of its possible applications’ (Ewald, 1991, p 198). Because economic, moral and political environments are always changing, insurance is always developing new ‘forms’, such as public liability insurance or house contents insurance. In their turn, these are the speciﬁc products of the ‘insurance imaginary’ – that is, ways of inventing useful purposes to which insurance technology may be put. Different forms of insurance are the specific ways in which the insurance
1 See also Simon, 1987; Knights and Verdubakis, 1993. Knights and Verdubakis (1993, p 739) point out that ‘[t]he relation then between actuarial calculations was not one of a straightforward application of mathematical formulae to insurance problems. Until halfway through the eighteenth century insurance premiums shared with gambling stakes, tontines and annuity rates a similar representation of risk based on a diversity of mathematical and non mathematical criteria’. See also Daston, 1988, p 240. However, Knights and Verdubakis move on more problematically to suggest that since that time insurance can analytically be deﬁned in terms of actuarial technologies.