![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
The central place occupied by actuarial calculation in insurance is usually understood as resulting from the process of bringing the laws of large numbers to bear on archival data in terms of an ‘insurance imaginary’. However, little attention is paid to the political and broader governmental conditions upon which actuarialism rests. Analysis of fire insurance in Australia indicates that despite urgings from other branches of insurance, the industry did not go down this ‘scientific’ track until well into the twentieth century. Instead, it relied on detailed individual inspection by insurance agents and a process of ‘cumulative dangerousness’ – adding up the multiple hazards discovered in each case and using this as a guide to setting premiums. The result was slow, expensive and cumbersome. But the industry was forced to adopt this approach because the built environment in Australian cities was so underregulated that nothing could be taken for granted. Only when adequately designed and enforced government regulations began to appear after the mid-1920s could fire insurance adopt actuarial techniques for economising uncertainty.
ACKNOWLEDGEMENTS
This research was funded by an Australian Research Committee Discovery Grant Project: DP1093187 ARC. We would like to thank Luis Lobo-Guerrero, Mariana Valverde, and two anonymous referees for the Journal of Cultural Economy for helpful comments on an earlier draft.
Notes
1. Thus we find the president of the Insurance Institute of Liverpool (UK) stating in his inaugural address (Alcock Citation1907, p. 310) that the most important feature of development in British insurance over the previous 40 years had been the ‘progress which fire (insurance) business has made toward becoming an exact science’. Whereas once things were ‘on a primitive basis’, now things had changed, and ‘there are now data to work upon of a nature sufficiently precise to satisfy the Institute or Faculty of Actuaries’.
2. Apart from the ‘normal’ fires, both Melbourne and Sydney suffered major multi-building conflagrations in their respective CBDs in this period. In Melbourne, the infamous ‘Block Fire” of November 1897 destroyed most of a large city block and gave rise to an inquiry into the adequacy of fire prevention and building regulations (Report of the Fire Prevention Inquiry Board 1898). Apart from some impact on fire brigades, the only major development was to limit the height to which buildings could be erected – resulting in a low rise urbanscape until the second half of the twentieth century (Schrader Citation2010). It was a conflagration in Pitt Street in central Sydney in 1890 that eventually gave rise to the formation of the Fire Underwriters Association (Cowper Citation1890, p. 684). A similar fire in Sydney 10 years later gave rise to a NSW Legislative Council (Citation1908) select committee on fire prevention in with similar lack of impact in that city.
3. For example, this letter is from the head office of Alliance Assurance to its Melbourne Office illustrates not only the level of surveillance but also the loose estimation that formed the basis of calculation of risks:
We would prefer not to hold more than we are at present retaining on the portion of the Twine Factory to which you refer, notwithstanding the cut-off by double iron doors from the part where carding is done, as we have not much faith in this form of separation of risk (Alliance Assurance Company Citation1909, 17 September). Monica Keneley (Citation2002) gives an account of the powerful British influence on Australian insurance after the 1890s depression.
4. See also the editorial of the Australasian Insurance and Banking Record (Citation1889, p. 595) which drew unfavourable comparisons with life insurance and urged that ‘there can be no doubt that the best results are obtained where experience is brought to bear upon statistics intelligently compiled’. Other unfavourable comparisons of this nature can be seen at Australasian Insurance and Banking Record 1908 (May) 436.
5. The ‘XYZ system’ referred to is the calculation of ‘risk’ by the addition of multiple factors relating to each individual risk – what is referred to in this paper as the ‘cumulative dangerousness’ model. Towns were classified A,B,C, and so forth; types of buildings – such as warehouses, post offices – were classified W,X,Y,Z, and so forth. These formed part of the ‘basis rates’, which were one of eight kinds of classificatory variable. Others included construction of building (degrees of) separation from other buildings, district in which located, additions to basis rates (such as the construction of neighbouring buildings), classification of occupation of the building, deductions for improvements and (most significantly) special loadings for specific hazards (Parrish Citation1915, pp. 5–6). Each variable in turn could be quite complex. As a result, of course, no two risks were identical.
6. The lack of access to actuarial data meant that the character and the skill of the insurance agent was also a critical matter to the insurance industry. Insurance agents wrote policies, inspected risks and reported on fire loss. The agents always had to send policies into the local fire office to be processed and approved. Complex matters were sent to London for approval where (as frequently was the case) the insurance company was British. The agents were carefully selected by the companies, and many were appointed on the basis of reports from local bank managers. Agents were assessed for their standing in the community, ability to assess moral hazards, length of residence in the community, and an assessment of their ability to generate new business. Nearly always, agents had other employment. Typical is this report on one David Christie: ‘He is a very old resident and much respected; holding beside other positions the secretaryships of the Hospital and local Agricultural Society. His office is in the best position and he should do fair business. The Office can thoroughly rely this agent; his knowledge of the town and district will enable him to make a good selection of risks’ (Royal Insurance Co. Citation1891–1908, 18 July 1892).
7. Tow is the name given to short flax fibres produced in the process of combing flax straw; flax and tow are both used in forms of linen production.
8. As this indicates, throughout the nineteenth century in Australia, such questions of moral hazard (the subjective characteristics of the insured party that increases the probability of loss to the insurer) were a central focus of the insurance agents’ evaluations of whether or not to insure. In the USA, Baranoff (Citation2003, p. 124) suggests that in the 1870s, the reputation of the applicant for insurance took precedence over the merits of the property in the decision to insure. In Britain, around the same time, Pearson (Citation2002, p. 7) has suggested this intense focus occurred because the fire insurance market was one in which ‘costs and benefits could not be precisely known’, precisely because the industry had not yet produced a body of scientific data that could be deployed in risk calculation. As a result, moral hazards could neither be ignored or tolerated, but neither could they be ‘fractured into economic factors’. To put it another way, ‘non-rational approaches to the problem of moral hazard’ were necessitated because of ‘random attitudes to the analysis of physical risk’ (Pearson Citation2002, p. 34). As a consequence, it could be assumed that the emergence of scientifically and actuarially based fire insurance would affect the place of moral hazard. Ericson and Doyle (Citation2004, p. 207) certainly appear to confirm Pearsons’ argument, noting that by the twentieth century, moral hazards were dealt with as risks to be spread, like any other economic factor. The effect, however, was not to eliminate moral hazard as an issue but to shift its place in the insurance process. Pressure was taken off the role of the insurance selling agent as a judge of moral character. In the twentieth century, moral hazard became far more a problem dealt with by loss adjustors. In passing, it is interesting to note that Grove Smith (Citation1918) in his monumental study of ‘fire waste’ in Canada, demanded that not only arson but also negligent actions leading to fires should be criminalised. He regarded this ‘human factor’ as the last truly ungoverned problem in the economics of urban fires. It may be an index of how far things had moved in ‘factoring-in’ moral hazard that this demand fell on completely deaf ears.
9. For a similar account see Ferguson (Citation1889, p. 784).
10. In practice things were not always so grounded and empirical. A South Australian insurance expert suggested that in addition to such minute inspection,
I desire to record my belief in what is known as ‘insurance instinct’. By this I mean the dislike an expert will sometimes feel to a risk, before or after survey, which cannot be traced to any distinct cause. I am of the opinion that a risk should never be accepted when this feeling exists, and I think I am justified by experience. (Tapley Citation1889, p. 878)
11. Goads was the most widely known publisher of fire plans in Britain and Australia, Sandborn took leading place in the USA and Canada. For a very useful and succinct account of fire maps and of Goads in particular, see Rowley (Citationn.d.).
12. See, for example, Woodward (Citation1979) for a North American history of these maps, and examples – which took basically the same format in Australia. For Australian examples, see Insurance Planning Association (Citation1923–28).
13. Even such an acknowledged expert on fire insurance rating as Parrish referred to it as ‘so intricate and involved that it is with feelings of trepidation I enter the portals’ ( Parrish Citation1915, p. 5).
14. Of course, this was not peculiar to the Australian context. As a contemporaneous American insurance writer noted, ‘The technical difficulties of estimating fire hazard, the diversity of risks, and the almost inevitable destruction of evidence of the cause of loss have combined to complicate the solution [of analysing hazard]’ (Quoted by Wermiel Citation2012, p. 239).
15. The ‘Tariff’ was a cartel of fire insurers created in 1897 and continuing in existence until the 1970s. Its primary role had originally been to reduce competition among fire insurers, and most especially to bring into line small insurers who offered coverage at uneconomic rates and were prone to collapse under the burden of a significant fire. The Fire Underwriters’ Association was the formal name given to the association at the heart of this arrangement with a brief to classify risks, set tariffs and maintain standards across the industry. Each state (until 1901, each colony) had its own association, but these were more or less standardised nationally. In many respects, the development of this organisation tracked that of its US counterpart (see Tebeau Citation2003).
16. The FUA of NSW had an ‘Interpretations and Revisions Committee’ that met weekly to consider suggested revisions to agreements, based on tests of fire appliances, materials, etc., or requests from insurance companies, municipal councils or manufacturers.
17. The NSW office of the FUA employed three surveyors, a sprinkler installation inspector and two electrical inspectors (FUA of NSW Committee Minutes 3 July 1928, p. 152. ML MSS 3155 Box H165).
18. For example, when in 1902, the Melbourne City Council carried out major upgrades to its electricity supply, the FUA inspector was dismayed at the work being carried out. Having condemned more than 10% of council-approved installations, he remarked:
I understand that the City Council Officers conduct tests of all the installations they connect to the supply mains, and that installations condemned by me have passed their test. The fact that an installation passes a fair installation test is no guarantee that it is a good one. It would be possible to get an excellent test with an installation wired with bare copper wire if it were properly installed, and yet no insurance rules or any other rules would allow bare copper to be used. (Report of the Electrical Inspector of the Fire Underwriters’ Association of Victoria, 24 April 1902)
19. We have been unable to find an unambiguous answer as to why government building and electrical regulation were so retarded. There is some evidence that this was regarded as a brake on development. Certainly, it had been a theme of opposition to building regulations as far back as the 1830s, which regarded regulation as a ‘cause of stagnation’ in the building industry (Scope Citation1837, p. 27). In 1919 and 1920, the Electrical Employers’ Association complained that enforcement of the Sydney Municipal Council's standards – regarded as woefully inadequate by insurers – had ‘caused a great deal of delay and inconvenience to users of electricity and annoyance to (its) members concerned’ (The Electrical Employers’ Association of NSW Citation1919, p. 2, 1920, pp. 2–3). Perhaps in response, the Council discontinued its practice of testing installations. In consequence, however, the association then complained that as a result cheap competition was installing inferior materials to a lower standard. It is possible this, in turn, gave impetus to the reforms of 1924–25.
20. At the fourth annual general meeting of the Fire Underwriters’ Association of Victoria, a letter was tabled from the association's acting electrical inspector who noted that during that year electrical supply for the city had been taken over by the City Council, and the Electric Light and Traction Co. Ltd. took over supply for the suburbs. He noted:
Within the next few months the current pressure will be raised for lighting from 100 to 200 volts, and for motor work from 100 to 400 volts. The increase in pressure will very materially increase the ‘short circuits’ and risk of Fire, and points to the necessity of being extra particular about the insulation of wires etc, which I will have to watch very closely, especially in connection with the renewal of old certificates. I would like to say that as a rule I do not have any difficulty in getting what I consider necessary carried out, though occasionally some fitter, through either ignorance of Electrical risks, or a desire to do their work cheaply, give me a good deal of trouble.
21. Prior to this, there had been a Victorian Inquiry into Fire Prevention in 1898 in the wake of a major conflagration in the central business district. The Inquiry Board found that regulations were inadequate, although much of its concern focused on the need to regulate the height of buildings, recommendations, nevertheless, were also made with respect to fire-resisting roofs, floors and ceilings, girders, doors, shutters and sprinklers. Of particular concern were provisions for the construction of stairs and lifts which were the focal concern of many fire experts in the late nineteenth century. However, while it was emphasised that in 1897, Melbourne still operated under a Building Act closely modelled on the Sydney Building Act of 1838, a new Building Act was not passed until 1928, although municipal councils received powers to make by-laws for building regulation and fire prevention in the Local Government Act 1915 (Vic.).
22. An almost exact parallel situation seems to have applied in South Australia. Ferguson (Citation1889, p. 783) noted that fire insurance offices owed much to the superintendent of Fire Brigades in Adelaide for his efforts to instil good fire preventive practices among tradespeople he inspected. With some bite, he remarked that the ‘superintendant, however, has no power to compel obedience to his instructions, and in many cases finds, on making a second call, that nothing has been done. If there be a municipal bye-law on this matter it should be rigidly enforced. If there be none our city fathers could profitably employ half an hour framing one.’ As well, he suggested that if an inspector were appointed to enforce such by-laws, then the penalties he might enforce during his first visit ‘would probably pay his first year's salary’.
23. Of course, insurers constantly bemoaned the lack of regulation. For example, a South Australian commentators stated that ‘civic authorities display the utmost apathy and indifference with regard to building regulations and the making of proper provision for fire extinction’ (Shepherd Citation1908, p. 309).
24. The state of Victoria generally was further advanced in these respects than were other states. The Local Government Act 1915 (Vic.) Section 198 gave municipal councils the power to make by-laws with respect to the construction of buildings, and in 1918, the Public Works Department provided model building by-laws. By 1919, the Victorian Public Works Department advised The Australian Insurance and Banking Record that ‘all but four metropolitan municipalities, inclusive of the City of South Melbourne, have adopted by laws under section 198 and they are fairly comprehensive’. The remaining municipalities appear to have fallen into line during the early 1920s. In NSW, the equivalent the Local Government Act 1919 (NSW) made municipal councils responsible for regulating building development in their area, although in the City of Sydney itself the process was to drag on for some years to come.
25. The result often is an ongoing struggle by insurers and states to externalise their respective costs and risks. Fire insurers are constant lobbyists for improved building standards, while governments complain that insurers do not do enough to ‘police’ their policyholders. Lobo-Guerrero (Citation2012, pp. 112–118) has examined an exactly parallel struggle with respect to the insurance of vessels against acts of modern piracy. The EU, for example, requires all ships flying under an EU flag to take out adequate insurance, on the basis of an expectation that insurers would insist on proper risk management against piracy. Insurers, on the other hand, complain that it is governments’ job to bring political security sufficient to allow their industry to function properly.
26. We wish to make clear that we are establishing the case for seeing state legislation and inspectorates as a condition of existence for actuarialism. That is, we see such legislation as a necessary but not sufficient condition for the change to occur in Australia. We are not therefore attempting to provide a causal account for the rise of state legislation in this area, as many other factors may have played a causal role. It is clear, as we have argued, that the FUA acted as a pressure group, and provided vital templates for government regulation. As in the USA, cartelisation was to play a major role in influencing state legislation (Baranoff Citation2005). In Canada, for example, O'Malley and Hutchinson (Citation2007a) have pointed to the role of increasing capital intensity in raising the interest of business corporations in fire prevention. It also is plausible that a key role was played by the rising importance of the relevant professions linked with the increasingly complex nature of urban material technologies. Changes in the insurance industry itself as it became perhaps the principal private financial formation in the Australian economy would have increased its political influence. The issues are many and complex. This question is the subject of a paper currently in preparation.