The Making of the Swedish Life Insurance Market 1855-1914
Liselotte Eriksson
ISSN: 1653-7475
Occasional Papers in Economic History
Licentiate Thesis
Occasional Papers in Economic History No. 15 2008
© Liselotte Eriksson, 2008
Institutionen för ekonomisk historia 901 87 Umeå
www.ekhist.umu.se
Abstract
This licentiate thesis examines the development of the life insurance industry during the period 1855-1914. The aim with the study is to recognise dimensions not frequently addressed by previous research on the insurance industry, namely the impact of social dimensions, including the implicit and explicit economic importance of social movements and the
diffusion of knowledge in society at large for the development of the life insurance industry.
The study shows that income and price had limited importance in explaining the demand for life insurance before the 20th century and that this can be attributed to a lack of sufficient knowledge regarding financial issues and to a far too high access cost in acquiring a life insurance for a large part of society. The development of the life insurance industry must therefore be understood through improved knowledge both on the part of the life insurance companies and on part of the consumers. The licentiate further shows how diffusion of knowledge throughout society also was due to a diffusion of democratic ideas and the rise of social movements, movements that life insurance actors were a part of. These actions helped open up the financial market for the masses and probably also strengthened the trust towards the industry. It is however hard to dismiss the life insurance actors’ engagement in women’s movement as a cover-up for other disguised motives not so honourable, while a direct economic gain for the life insurance industry is hard to establish.
Keywords: Economic history, life insurance, demand, price elasticity, income elasticity, married women’s property rights, social movement
Acknowledgements
Many have contributed to this licentiate thesis. First I wish to thank my supervisor Magnus Lindmark for his constant encouragement, entertaining metaphors and for commenting this licentiate thesis in various stages without loosing faith. I further want to thank several eminent academics for their often vivid comments and support: Mike Adams, Lars Fredrik Andersson, Monica Burman, Mikael Lönnborg, Joel Mokyr and Jan Ottosson.
On the personal level my twin sister Madeleine and her husband Rikard have my constant love and gratitude. And to my beloved niece Judith born October the 16th 2008, I am sorry I missed your first weeks on this earth while finishing this licentiate thesis. It won’t happen again. Finally, thank you Fredrik for everything!
Liselotte Eriksson
Umeå, October 31th 2008
CONTENTS
1 INTRODUCTION... 6
1.1BACKGROUND... 6
1.2THE OBJECTIVES AND RESEARCH QUESTIONS... 10
1.3THE OUTLINE OF THE LICENTIATE THESIS AND DEFINITIONS... 12
1.5METHODS AND SOURCES... 14
1.5.1 Article I... 16
1.5.2 Article II... 17
2. PREVIOUS RESEARCH ... 19
3. USEFUL KNOWLEDGE AND THE FINANCIAL ECONOMY ... 25
ARTICLE 1: INSURANCE AND INCOME GROWTH: THE CASE OF SWEDEN 1830-1950 ... 28
Abstract ... 30
Key Words ... 30
Introduction ... 31
The financial revolution and the growth of insurance markets ... 33
Method... 36
Data ... 38
The demand for insurance ... 43
The development of price and risk in life and fire insurance... 50
Concluding remarks ... 57
Acknowledgements ... 59
Literature... 60
ARTICLE 2: FINANCIAL BUSINESS AS A SOCIAL PROJECT... 67
6 CONCLUSIONS ... 91
7 SOURCES AND LITERATURE... 93
7.1OFFICIAL STATISTICS... 93
7.2LITERATURE... 93
1 Introduction
1.1 Background
Sound insurance companies, are according to Richard Sylla one of the components of a modern financial system. The evolution of sound insurance companies is part of what he calls the 19th century financial revolution.1 Not the least the life insurance industry came to play a crucial part in the development of the financial system by mobilizing savings and allocating risks.2
The social and economic transformation of Sweden during the 19th century changed how risks were perceived, valued and mitigated. In the agricultural society, it was commonplace that risk-management was based on customary rules that aimed at diversifying risks in the agricultural production, i.e. self-insurance.3 Women, children or other relatives could for example have opportunity and enough knowledge about e.g. farming to take the place of a deceased or injured member of the family. Urbanisation and new conditions on the labour market changed this state of affairs as wage earnings became increasingly important. An increasing part of the population became dependent primarily on one male breadwinner, not the least in the growing cities. The premature death of the man could therefore leave the family in economic ruin. A life insurance could in these cases be a substitute for the lack of capital assets and the over time diminishing safety net of the old agricultural society.
Consequently, life insurance came to play an important part of the 19th century economic and social modernisation.
Organized efforts to reduce the consequences of private risk, by sharing risk, have a long tradition in society. Already medieval guilds and craftsmen’s associations initiated a primitive kind of compulsory health and funeral insurance, and in the case of illness, everyone in the guild were obligated to support the injured or seriously ill guild-member by economic or practical means. The receiver of the contributions, however, had to pay back the help given after his recovery. In the case of death, all guild-members had to make a contribution to the funeral.4 This system had its weaknesses, for instance due to insufficient funding and risk management while, from the perspective of the individual, a period of illness could result in serious indebtedness. A forerunner to the life insurance industry as it evolved in Sweden since
1 Sylla (2002), p. 280
2 Mason (1995), pp. 172-175.
3 Hägg (1998), p. 106.
4 Lindberg (1949), pp. 2-5.
the mid 19th century, was a kind of widows’ fund (Amiralitets-Krigsmanskassan), established in Sweden 1642. This was, in the Swedish context, a new institutional form of provision for dependents, providing pensions for widows on the basis of contributions made during their husbands’ lifetimes.5 The funds were limited to specific regions and underwriting was exclusively intended for men with specific positions in society, primarily the growing group of officials with lack of capital assets, mainly priests and soldiery. The idea underlying the widows’ funds was to avoid a degeneration of upper- and bourgeois widows by enabling them to keep the same standard even after the death of the husband. The worst case scenario would be if widows of officials were forced to work to make a living and by that became part of the working class. This would affect the remaining family of the widow and shake the legitimacy of the class-society. According to Rosenhaft, the upper-classes therefore joined in effort to avoid a similar development and widows’ funds became widely used. Rosenhaft suggests that the demand for life insurance was generated from the demand of women. According to Rosenhaft widows’ funds being a forerunner to life insurance, therefore implies that women invented life insurance. 6
Over time different kinds of health- and burial insurance societies (Sjuk- and
begravningskassor) evolved, many contemporary with the modern insurance industry.7 These societies were however not nationwide but limited either to a working place, unions, a group like total abstainers, a parish or a city. In Sweden, these societies gained public financial support by different arrangements settled during the years 1891-1910, if they adjusted their business to certain conditions. The state wished to enhance the control and sound operations of these funds, since both the widows’ funds and several of the earlier insurance societies, lacked sufficient actuarial knowledge. All age groups could for example pay the same premiums, which lead to increased risk of insolvency, leaving people without economic protection and without the right to claim the premiums already paid.8 Despite this, the societies could in some cases grow strong, while they often had a local connection and was run by people known to all members of the society. This could, according to several scholars, lead to a misplaced trust in insolvent societies.9
It is evident that the guild system, the widows’ funds and the health- and burial societies, all based their business on the practise of excluding and including customers by
5 Three different widows’ funds were founded in Sweden during the second half of the 18th century.
6 Rosenhaft (2004), p. 163.
7 The societies were of limited importance until 1870, according to Lindeberg (1949), p. 52.
8 Bergander (1967), p. 285.
9 Hägg (1998), pp. 217-218. See also Stalson (1942), pp. 35-36.
guild, class, occupation, sex or geographic belonging. This was according to Hägg an advantage, while trust in the society already was established by the mutual organisation, the familiarity and recognition of the organizers of the societies.10
Hence, this licentiate thesis will argue that with the arrival of the life insurance companies, the formal excluding vanished and a safety net in shape of a life insurance contract was, at least in theory, offered to everyone who could afford. This liberal idea was however in practice encompassing the few. Like the widows funds, the purchase of a life insurance, initially, to a high extent was limited to bourgeois or upper-class male officials, women and blue-collar workers among the insured were quite unusual. The life insurance was early on both expensive and inaccessible, especially for people living in remote areas of Sweden.
Conditions put forward explaining the establishment of the first national insurance company based on robust mortality tables and actuarial techniques, was among other factors affected by freedom of trade through the new statues of 1846 and the abolishment of the medieval, compulsory public fire aid in 1853.11 This compulsory fire aid implied that all members of a county or parish were forced to make contributions to victims of fire damages.
This risk-sharing institution was established in provincial laws already in 1200 and in the General Rural Law hundred-and-fifty years later, it did however not encompass urban areas, where instead local mutual fire insurance companies were introduced in the 18th century.12
Initially a state owned life insurance company was planned in Sweden, with several German annuity assurance companies as role models. But the private interests preceded the plans, and the composite life- and fire insurance company Skandia was established 1855.13 Influential, wealthy men with political influence were supporters of the private initiative, further illustrated by members of parliament becoming economic supporters and board members in Skandia.14
The historic trend of the corporate forms in Swedish life insurance industry can be characterized by two waves of establishment: The first companies that were established were composite joint-stock companies. 15 The composite companies, with both life- and fire
10 Hägg (1998), pp. 215-218.
11 Bucht (1936), pp. 136-168.
12 Söderberg (1935), pp. 21-31. See also Larsson & Lönnborg (2008, forthcoming), pp. 3-4.
13 Following the corporate structure of Skandia, Svea was established in 1866 in Gothenburg, Skåne in 1884 in Malmö. The first company limited to life insurance only, was Nordstjernan established in 1871.
14 For example A.O. Wallenberg.
15 The joint-stock company Thule est. 1875 introduced mutual principles with e.g. a small capital stock. It also limited the exchange for shareholders in addition to the one for insurers and the costumers had the symbolic
insurance were adopted in an era where fire insurance was measured as a much lower risk than life insurance, probably because fire insurance was an older, experienced line of
insurance. This however, turned out to be a false assumption clearly displayed by the city fires of the late nineteenth century and the decline in mortality.16 During the second half of the 19th century new insurance companies adapted the mutual idea and a number of small mutual life insurance companies were established; by 1890 twenty mutual life insurance companies were active. But due to, among other things, insufficient risk-calculations, only eight (widows’
funds excluded) of the twenty mutual companies survived up till the turn of the century.17 By this time, a new kind of mutual life insurance, the industrial life insurance
(folkförsäkringsbolag), started to successfully insure the working class. This development was related to transformations occurring on different social and economic levels in society, for instance, the insurance companies recognized an unexploited market in the working class. The high risks faced by the working class, lacking both protection in case of injury or death, opened up for the industrial life insurance companies.
By aiming their business towards the working classes, the business of the mutual companies did further democratize –in the sense underwriting larger and larger shares of the population- the insurance industry by adjusting its products to working class conditions. The business idea of the industrial life insurance was in other words to offer smaller life
insurances something that demanded a more extensive organisation of insurance agents to collect the often weekly or monthly premiums paid by the household of the insured.
The focus of this licentiate thesis is however the early ordinary life insurance, which was aimed primarily towards the middle-class. The middle-class was an important target group for the life insurance industry, while the middle class often were dependent on wage earnings, and not necessarily owners of capital assets; a life insurance could therefore offer the protection needed. Previously in this introduction it was stated that life insurance became a crucial mean to secure the family income after the death of the breadwinner. Furthermore, it remains obvious that, by doing so, the life insurance also facilitated the growth of the middle class. The liberal middle class in turn became important while they initiated and added fuel to the growing debates on social and economic justice during the nineteenth century. The
implication of life insurance is therefore that it facilitated and made possible the conditions that gave rise to social movements with intent to loosen up notions of suppression that still
possibility to elect new managers. Variations of this company structure came to set the trend, see Larsson (1998), p. 53.
16 Bergander (1962), pp. 71-72
17 Englund (1982), p. 46.
permeated society during the second half of the 19th century. By doing so, the life insurance also challenged the traditional notions of reciprocity of the pre-industrial society. Life insurance therefore came in the centre of 19th century economic, social and economic transformation.
1.2 The objectives and research questions
In the international literature, several researches have noticed a time-lag between the establishment of a financial industry on one hand, and the maturity and diffusion of a financial service on the other. This has also been stressed by Sylla: “The British financial system coming out of the early-eighteenth-century financial revolution did not really mature into a fully modern system until sometime in the nineteenth century.”18 This reasoning can also be applied on the Swedish life insurance industry where the first company was
established 1855. By 1898 there were approximately 120 000 persons insured in Swedish life insurance companies, but ten years later,1908, this number had reached to closely one
million.19
It might be argued that every new industry has problems when establishing its business, but a number of scholars on the insurance industry consider this problem to be especially accentuated in the establishment of the life insurance industry, even when compared to other financial services. 20
One reason for this is that the social mechanisms and the behaviour of ordinary people is crucial for understanding the pattern of demand for insurance, while the acting and choices of people in society, together with other economic factors, creates the conditions for the market, figuratively speaking the rules of the game. Still previous Swedish research on the insurance industry has so far not been including social history in explaining this process, although both economic and social factors are discussed in the international literature. It is therefore reason to hypothesise that the interplay between economic and social factors, among other things, constituted the qualitative difference between the developments prior to the 1850s and the broad financial break-thorough characterizing 19th century Sweden.21
18 Sylla (2002), p. 281.
19 Civildepartementet försäkringsinspektionen, Försäkringsväsendet i riket1898-1908, Stockholm.
20 See Stalson (1942), pp. 37-44. Zelizer (1979), pp.149-153. Lönnborg (1999), p. 62. Kenely (2001), p. 165 and Wright & Smith (2004), p. 3.
21The Swedish nodes in the traditional financial network were among others the numerous urban dealers and brokers, together with prosperous farmers and peasants which actively sought outincome-generating
investment opportunities according to Lindgren (2002), p. 811.
The importance of diffusion of technological ideas in explaining economic development has been emphasised by several scholars.22 From the perspective of life insurance, technological ideas would pertain to both actuarial techniques and financial risk management.
However, Mokyr furthermore stresses the importance of society at large and its improved access to different ideas with potential of leading to social progress, in explaining the economic development of the industrial society.23 With respect to insurance this means, broadly speaking that, what the general society “knows”, is crucial for the development of the insurance industry, in other words the general public’s knowledge of life insurance as an economic instrument for risk sharing and risk mitigation.
The overall purpose of this licentiate thesis is therefore to explore how the transition of the Swedish life insurance industry into a matured financial system, interacted with socio- economic and political change during the period 1855-1914. This is motivated since aspects on the Swedish financial development this far have been focusing primarily on the specific businesses relationships, and the specific regulation of the industry as a whole, while the co- evolution of financial modernisation and social improvement and social change has been given much less attention. The economic and social dimensions of modernisation and economic growth can hardly be separated.24 According to Mokyr the economic development and improvement that took place during the 19th century economic expansion, also involved social improvement and social change coming out of social networks, institutions and organisations of different kind.
This main purpose is divided into two objectives for a closer study of the determinants of demand and how social transformation and the development of life insurance industry reciprocated.
- The first objective is to identify the existence, proper periodisation and nature of the process of maturity of the Swedish life insurance industry.
- The second objective is to understand how the process of maturity of the life insurance industry was influenced and how it influenced social and political change, alas exclusive of factors directly associated with economic growth.
22See for example Landes (1969). Rosenberg (1982) and Bruland (1998).
23 Mokyr (2002), pp. 8, 32, 37.
24 Mokyr (2005), p. 291.
Aspects of these objectives are then studied in two articles: (1) The first article analyses the demand for life insurance, and (2) The second article explores, among others, the role of insurance managers in the first organisation for women’s rights, which in turn is related to property rights issues.
The first objective is dealt with in the first article which seeks to identify the process of maturity through studying the impact of traditional demand-side factors on the historical demand for life insurance. By doing so, factors related to the maturity of the industry, such as the diffusion of useful knowledge may be interpreted and discussed qualitatively. Here, a theoretical framework based on the ideas of Joel Mokyr will be used.
More specifically the research question is “when did the insurance industry mature in Sweden and what factors may explain this”?
The second objective is met by an analysis of the roles played by social and political movements in the development of the life insurance industry, but also vice versa: how the life insurance industry affected social and political movements. This implies investigate the period which turned Sweden into a financially modernized country and almost simultaneously gave rise to extended rights for women and blue-collar workers.
More specifically the research question is, are there evidence for insurance actor involvement in social and political movements and, if that is the case, how tightly connected was their motives to those of the companies and their owners?
Previous research has studied the transformation of Swedish private financial markets towards more institutionalised forms of financial institutions.25 The concern of this paper is, however, to study the process that took place within an already modernised life insurance industry, from the establishment of the industry and towards the maturity of life insurance, which is here seen as the evolution towards the establishment of the concept of life insurance in people’s everyday life.
1.3 The outline of the licentiate thesis and definitions
In the preceding section of this study the term “modern financial system” has already been used, which requires a definition. This study is going to use the definition made by Sylla who defines the components of a modern financial system to be: “sound public finances and public debt management; stable monetary and payments arrangements; sound banking systems
25 Lilja (2004), pp. 42-45. Shows how the private credit market in Sweden existed before and during the transformation to a more institutionalised credit market.
(more generally, institutional lenders); an effective central bank; good securities markets for debt, equity, and money-market instruments; and sound insurance companies (more generally, institutional investors).”26 According to Ögren these criteria got realised for the Swedish financial system around 1850-1870.27 This underlines what already has been stated, that the life insurance companies28 included in this study were, according to Sylla and Ögren, part of a modern financial system.
This licentiate thesis is further focusing on the establishment of the modern life insurance industry up until the “maturity” of the same. Sylla is also quoted in defining the
“maturity” of the life insurance industry. He refers to a pre-mature stage where the financial system was indeed modern but was lacking the incentives to develop the economy.29 Of course it is hard to define when a modern financial service turns “mature”. This licentiate thesis is however defining maturity as a quantitative and qualitative difference taking place within the industry, making possible a broad public use of life insurance. It is quite
instrumentally defined as the point in time where demand for life insurance converges with the income elasticity of demand observed in the beginning of the 20th century.
The licentiate thesis encompasses roughly the period from 1855-1914, but in some cases years before and after the period have been of interest and included in the study. The initiating year is the year of establishment of the first Swedish life insurance company and the closing year of the study is chiefly motivated by the introduction of the public pension system in Sweden 1913, which changed the conditions of the private life insurance market.30 The impact of the First World War did of course also change the state of affairs for the industry.
This study is furthermore focusing on the Swedish development. By investigating the private insurance industry the immediate period before the expansion of the Swedish welfare state, this study offers a link between private insurance and the forthcoming system of public insurance characterizing the internationally well known Swedish welfare state.
26 Sylla (2002), p. 280.
27 Ögren (2005), pp. 64-93.
28 With the term “life insurance industry”, signifies all insurance companies that carried life insurance, which is:
joint-stock companies, mutual companies and composite companies – companies carrying both fire- and life insurance.
29 Sylla (2002), p. 281.
30 Englund (1982), pp. 60-61, argues the most important change for the insurance industry in modern times to be the expansion of the public insurance system. Here, he only includes the impact of the public accident insurance on the private and the competition brought about by the state owned insurance institution
Riksförsäkringsanstalten. There are some evidence however that the public pension system made it harder for insurance agents to sell private life insurance see Borgh (1913), p. 7.
1.5 Methods and sources
While this licentiate thesis comprises of two articles and a compounding text, a general overview regarding methods and sources is first outlined. Subsequently an in-depth analysis of methods and sources regarding the two articles is presented.
There is a wide range of historical sources on the subject of insurance. The insurance industry has left behind books in the shape of company monographs; and journals, like the insurance industry journals Försäkringsföreningens tidskrift and Gjallarhornet. In this
licentiate thesis, the insurance industry journal has been of major interest partly because of its statistical content and partly because of a specific interest in the members of the Swedish Insurance Society. Furthermore, proceedings of the insurance companies are preserved in archives. There is also a rich set of official statistics on the insurance industry and official parliamentary publications concerning regulations of the insurance industry.
Insurance monographs have had a great importance in the making of this licentiate thesis and the writing of the same has a long tradition in Sweden. Almost every Swedish insurance company that ever existed, even if it were just for a few years, has their own monograph. The monographs on the other hand vary in extensiveness and quality and hold both different insurance statistics and discussions of values and goals of the company besides discussions of the society at large. Different monographs can have very different content and of small, short lived insurance companies, like Oden and Valand, there only remains a brief monograph more like a pamphlet, while the successful, large insurance companies like Skandia and Thule have produced several monographs with extensive statistical content. A to heavy utilisation of insurance monographs might therefore lead to an unbalanced picture of the development of the insurance industry, while it is generally acknowledged that the winners write the history.
Insurance monographs also bring up the question of the dividing line between research and sources. For the sake of simplicity, all early insurance monographs are considered secondary sources in this study; but that is not to say that many insurance monographs not have a high qualitative content, like company monographs as: Lifförsäkrings-aktiebolaget Thule 1873-1898: festskrift vid bolagets tjugofemårsjubileum, I and II; Brand- and
aktiebolaget Svea 1866-1916; Försäkringsaktiebolaget Skandia 1855-1930 I, II and Svenska lifförsäkringsanstalten Trygg 1899-1909.
This selection rule means that modern insurance monographs like the ones on Skandia by Englund from 1982 and Kuuse and Olsson from 2002 are considered as previous research,
while there is a considerable qualitative difference between Englunds, Kuuse and Olssons work and several early monographs of insurance companies.31 Some early monographs must be considered tendentious and used with caution, especially relating to the company’s own description of events and conflicts.
The insurance industry also documented their accomplishment collectively in journals and periodicals. The insurance industry journal documented the actions of the Swedish Insurance Society (Försäkringsföreningen) during the period 1878-1920, 1921 the journal changed its name to Nordic Insurance Journal (Nordisk Försäkrings Tidskrift). The journal is a rich source of historical data and contains annual statistics, like balance sheets of the companies. The journals however, as well possess a broad qualitative content. They give a valuable insight to the contemporary problems of the companies; to topics of interest and of conflict, by documenting meetings, discussions and lectures in the insurance organisation.
Large volumes of historical material of the insurance industry have also been preserved in archives, namely in the Centre of trade and business history (Centrum för
näringslivshistoria). Because of the many fusions of insurance companies, the proceedings of merged companies and bankrupt companies are gathered under two archival creators: Skandia and Scandinavian Private Bank (SEB). The Skandia archive primarily retains information regarding the joint-stock companies Skandia, Thule, Nordstjernan, Skåne and Svea. While SEB holds Trygg, Victoria and De Förenade. The archives keep a wide range of material, like board records and insurance registers. The material of some small and short-lived companies like Valand and Balder are however for ever lost. The archive has been visited, but because of the nature of the questions of this licentiate thesis, material of the archive has not been used, while the insurance industry journal was found to be the most informative source.32
From the year 1889-1903 the ministry of public administration also collected insurance statistics, more all-embracing then the one collected by the insurance industry journal.
When the insurance industry got regulated by the state in 1903, the Swedish private insurance supervisory service (Försäkringsinspektionen) requested and compiled insurance statistics between the years 1903-1912.33 These data series was thereafter followed by Private
31 Kuuse & Olsson (2000) and Englund (1982).
32 Material of CfN is going to be used in the forthcoming articles.
33 Försäkringsinspektionen, Försäkringsinspektionens underdåniga berättelse till Kungl. Maj:t beträffande försäkringsväsendet i riket år 1903-1912, Stockholm.
insurance companies (Enskilda försäkringsanstalter) also published by the Swedish private insurance supervisory service 1914-1984.34
Insurance data has with other words been compiled by different organisations and authorities; this has unfortunately resulted in some incomplete series. One example is that data on insured persons ceased in 1911.
Historical sources of the insurance industry also include official parliamentary
publications. The main goal of the insurance business own organisation was to work towards an insurance legislation, and several insurance actors were also members of Parliament. The organisation also functioned as a consultation body.
As illustrated above, the available sources are comprehensive and are well suited for engaging in both quantitative and qualitative research. The existing statistical material is of a considerable size and large parts has not yet been collected and analysed in previous research, the data, with other words, still holds the answer to a variety of different research questions of international interest.
The belief is that the qualities of quantitative and qualitative methods combined give an enhanced explanatory power to this licentiate thesis, by study parts of the same phenomena but from two different perspectives and on two different levels.
The methodological approach and evaluation of the sources regarding the specific articles are as follows:
1.5.1 Article I
The first article deals with the elasticity of demand in life and fire insurance during the period 1850 to 1950. In order to avoid certain difficulties when deflating consumer expenditure data from various household budget investigations, the share of food stuff is used as an indicator of the consumption space of the households. To estimate the elasticity in fire and life insurance we have used a two-step model based on data from household budget surveys and time series data on insurance premiums and consumption. First, the elasticity coefficients are estimated by an OLS-model on the household budget data, seven cross-sections between the years 1913- 1978.
Secondly, the estimated coefficients are employed to back-cast household expenses on fire and life insurance during the period 1850 to 1950. To avoid dealing with aspects of
34 SOS, Enskilda försäkringsanstalter1914-1984 (Swedish Official Statistics Private Insurance), Försäkringsinspektionen, Stockholm.
deflation concerning changes in relative prices during the period of study, the back-cast is, as previously mentioned, based on food share. The estimated series on expenses are finally compared with the actual premium incomes derived from the Swedish life and fire insurance industry during the same period.
The method of constructing time series based on cross-section estimations might imply methodological problems. The first obvious dilemma is the lack of cross-section estimations before 1913, which brings up a discussion regarding counterfactual methods. Counterfactual methods can sometimes be considered a-historic, by applying today’s notions on historical data by employing the pattern of demand from one period into another. The point of departure of the article is however not, that the pattern of demand had the same components throughout history, but to conclude that there were differences and instead focusing on a discussion how these differences can be explained.
Other data, as data on premium and claims is derived from Bergander, the insurance industry journal and official statistics.When data has been unavailable, companies own collected data in monographs has been used. This has primarily been employed regarding the fire insurance industry established before or early in the twentieth century.
In dealing with historical data, the data is very rarely complete. When it comes to data missing occasional years, additional estimates have been made. Assumptions have also been made in maintaining data concerning the foreign insurance business in Sweden.
1.5.2 Article II
The primary material in article two consists of official parliamentary publications, reports from the organisation for married property rights, the Home Journal (Tidskrift för hemmet renamed 1886 to Dagny) and the insurance industry journal (Försäkringsföreningens tidskrift).
Moving first to the Parliamentary debates, as a selection rule, the bills selected and analysed was to be directly connected to the married women’s property act of 1874. All bills on married women’s property rights could not be included in this study considering the
limitations put on an article. This selection rule resulted in three bills, one bill in 1873 and two bills in 1874. While the debates after the bills of 1874 are brief, the focus is mainly on the debate preceding the bill of 1873, a debate that lasted several hours and attracted general attention.
For illustrating the forthcoming political mentality and the continuously social engagement of life insurance actors, all parliamentary publications 1875 until 1920 are considered. The principles for the bills selected after 1874 are therefore solely that they illustrate the engagement of life insurance actors in women’s rights issues.
The last bill from 1895 was selected while the life insurance industry was directly involved in the proposal. Moreover, the sources on the insurance industry are also very much appropriate for qualitative research. The monographs and journals are apposite for analysing the ideological values and goals of each company and the whole insurance industry combined.
The method of analysing Parliamentary debates might in some cases be problematic while the voting and acting of Parliamentary members not always shows their true opinions.
The insurance industry journal instead gives a unique possibility to study debates and arguments concerning a specific question before it were put forward in Parliament, this strengthens the reliability of the analysis.
Other primary material used is the Home Journal and Dagny, which functioned as a spokesman for women’s movement. The organisation for married women’s property rights had their own report series, which were published in the home journal. This material is used to include how women’s movements were working and debating concerning the bills. The material from the home journal covers the period 1873-1895 and the number of volumes per year varies between six and four. The report series of the organisation for married women’s property rights encompasses the years 1874 until the discontinuance of the organisation in 1895.
For the purpose of studying debates among insurance actors concerning issues related to married women’s property rights and laws concerning insurance contracts, the annual
insurance journals are used. The first time the topics were up for discussion was 1884 and the last time 1906. Discussions regarding insurance contracts were by all means taking place after 1906 but without the dimension of improving for married women.
A scope of this study is the identification of life insurance actors in the organisation for married women’s property rights. In doing so unpublished registers of members in the organisation has been used. These registers are handwritten by each member signing her or his own name. Some of these signatures have been hard or impossible to make out. But the objective of the article is not to map all members, the high representation of insurance actors must in either case be considered established.
In further identifying the actors, registers in the insurance journal, over board members and agents, also have been used. To establish the life insurance actors’ social networks and
engagement in other social movements, biographies35, literature on social movements36 and reference books have been employed.37
This licentiate thesis does not however deny that there were other men than life insurance actors, engaged in women’s rights during the period of study. The focus is on life insurance actors and therefore the material selected is not representative for the regular Parliamentary proceedings during this time. Material of this type also primarily describes the makings of later or contemporary famous persons while the infamous but equal hard working passes unnoticed.
The primary identities of some actors of this study were probably scientist,
entrepreneurs, politicians or estate owner’s rather then life insurance actors. With other words they could be board members in life insurance companies on the side. But several of them obviously were willing to risk their capital in life insurance companies.
2. Previous research
One of the most debated questions in international research on life insurance is why the demand for life insurance seemed to have been lower compared to other types of insurance.38 Researches claim observing a gap between the establishment of the life insurance industry and its take-off or maturity. This phenomenon has been studied from several theoretical
perspectives. The most common assumption is that the life insurance business would have been affected by the dynamics of growing incomes and urbanisation.
The importance of income elasticity of demand in explaining the growth of insurance has been stressed by a number of scholars such as Keneley, in the case of the Australian life insurance market during the 1950s and 1960s,39 Beenstock, Dickinson and Khajura, where they argue that insurance activity is highly dependent on the level of domestic per capita income.40 Despite the number of demand side factors put forward, previous research has not been concerned with explicit testing of demand factors.
35 See Englund (1993) and Weidel Randver (1981).
36 See Skarstedt (1903). Linder (1918) and Bolin (1930).
37 See Bohman (1942).
38 The majority of researches claim life insurance to be “a sold good, not a bought good.” See Stalson (1942);
Zelizer (1979) and Wright & Smith (2004). This notion has however also been challenged, Di Matteo & Emery argue that the life insurance was a sought asset, because of its role in the household portfolio during the course of the life cycle, Di Matteo & Emery (2002).
39 Kenely (2002), pp. 54-76.
40 Beenstock, Dickinson & Khajura (1988), pp. 259-272.
Non-quantitative factors have too been emphasized in explaining the demand for insurance and the slow development of the industry. To some researches the circumstances surrounding the life insurance seems contradictory.For example Wright and Smith claim life insurance should be the easiest insurance to sell, as everyone eventually dies. But not every house goes up in flames and not every ship sinks, still fire- and marine insurance got established on the market faster than life insurance both internationally and in Sweden.41
Zelizer claims demand for life insurance was hindered by moral and religious objections towards the industry and according to Stalson intensive marketing and an extensive system of insurance agents accounted for the increasing demand for life insurance in America during the 19th century. Murphy on the other hand addresses the significance of the changing social structures for the demand of life insurance in America, she consider the rise of the middle- class to be the driving force of demand for life insurance.42
Hägg argues the demand for regulation and supervision could be means of increasing the demand of consumers, while regulation fostered trust which was essential for the demand for trust-sensitive industries like insurance.43
Although social dimensions have been addressed, the direct interaction between the insurance industry and movements of social change has not been a matter of study, and Pearson and Richardson notes: “Few have attempted to examine the business population of a specific location in its entirety and systematically to analyse all its activities and economic, social, religious, political, cultural, and familial linkages.” They continue: “[…] there’s a lack of studies concerning collective diversification by fairly homogeneous business groups moving out of core trading or manufacturing activities into other areas”. The authors argue that such diversification was crucial to economic change. 44 This licentiate thesis further suggests that such diversification also was crucial to social change.
The question of the second article is therefore if there is evidence for insurance actor involvement in social and political movements and, if that is the case, how tightly connected was their motives to those of the companies and their owners?
Insurance actors recognized the problems for married women to claim the sum insured and the lack of regulations of insurance contracting. In answering the question of the second article, problems regarding insurance regulation becomes of interest.
41 Wright & Smith (2004), p. 21 and Lönnborg (1999), p. 62.
42 Murphy (2005), p. 262.
43 Hägg (1998), p. 82.
44 Pearson & Richardson (2002), p. 659.
Insurance regulation with focus on regulatory regimes and the role of institutions in the sense of insurance industry regulation are very well investigated areas in previous research on the insurance industry.
Bergander in Försäkringsväsendet i Sverige from 1962, presents several insurance regulations during the period 1886-1927.45 He gives among other things a comprehensive overview regarding the establishment of the co-operation between companies and the agreement of self-regulation by for instance the Swedish Insurance Society (Svenska försäkringsföreningen). The lack of regulation of the insurance industry during the 19th
century, made the industry itself create informal norms and rules of regulation. The insurance industry was practically self-regulated until 1903, while legislators tried to obtain a
harmonisation of the Swedish laws with the ones in Norway and Denmark. The Swedish insurance society was constituted by insurance executives and other insurance actors, in the beginning just from join-stock companies, with the purpose to work towards an insurance legislation; the insurance industry was with other words very much involved in the legislative process. One reason for the wish to formulate an insurance legislation was that the joint-stock companies complied with the joint-stock company law of 1848, and the lack of regulation for mutual companies lead to conflicts between joint-stock insurance companies and mutual companies.46 This was the explanation for the involvement of insurance actors in the making of the law of 1903, according to Englund in Försäkring och fusioner (Insurance and
fusions).47
In Den reglerade marknaden (The regulated market) from 1991, Larsson analyses the change and structural development of the insurance industry during the period 1850-1980.
Focus is on the relation between the state and the financial market and the insurance regulation acts of 1903, 1917, 1948, although other regulations by the state besides the insurance regulation are mentioned.48 Regulations and public insurance had varied impact on the private insurance industry. The public accident insurance of 1918, for example, implied that private accident insurance companies became discriminated by a direct intervention of the state in the establishment of new accident insurance companies.49 The common feature of
45 Bergander (1967), pp. 346-365.
46 Larsson (1991), p. 102.
47 Englund (1982), pp. 52-53.
48 Larsson (1991), p. 29.
49 Larsson (1991), p. 28.
the insurance regulations was that they had the purpose of protecting the insured, but also the established companies through limiting what was perceived as unsound competition.50
In An institutional analysis of insurance regulation, from 1998, Hägg argues that decentralised enforcement mechanisms like the Swedish Insurance Society, promoted honest trade and valued long lasting relations and a good reputation. In the lack of state regulation the industry shared experiences and monitored each other.51 The dynamics of the market forced managers to act as they had the interest of consumers at heart. The reason for the co- operating insurance industry to later find it urgent to gain trust-support by public regulation was that the young insurance market had to rely on repeated exchange and reputation to be viewed as trustworthy. And in this case the life insurance companies faced stricter regulations than non-life insurers according to Larsson and Lönnborg.52
Although Hägg does not study the law of insurance contracting empirically, he
theoretically argues that essential to insurance contracting are the notions of public trust and legal enforcement in the course of economic contracting.But the novelty in mixture with long-term contracts implied that trust in the insurer based on repetitive trade was not
established among the general public. Hägg therefore claims that if the seller cannot persuade potential buyers to buy insurance due to a too low level of trust, the insurer will seek a third party to give guarantees or punish deviations from commitments and by this create the trust needed for the development of the industry.53
The main incentive for insurance regulation could therefore be explained as an institutional negative response of the public towards the industry, and a wish to realise economic cost benefits from improved insurance contracting, and greater certainty of legal enforcement and property rights.54 Lönnborg states that a primary goal of the private actors to promote a regulation of the insurance market was to reduce competition. This is also
supported by Lindmark, Andersson and Adams who also claim that the early motivation for the Insurance Act of 1886 to be originated from the desire of insurance companies to promote their economic self-interests by tightening controls on market entry.55 But Hägg states
however that the Swedish Insurance Society cannot be rejected as a cover-up for disguised motives not so honourable.56 Lindmark et. al. further argue that regulations imposed
50 Larsson (1998), p. 129.
51 Hägg (1998), pp. 186-188.
52 Larsson & Lönnborg (2008, forthcoming), p. 10.
53 Hägg (1998), pp. 215-218.
54 Hägg (1998), pp. 48-49.
55 Lönnborg (1999), p. 82. Lindmark, Andersson & Adams (2006), pp. 362-363.
56 Hägg (1998), p. 188.
subsequently by legislation could be interpreted as reflecting the will of the public
administration to enhance its political power during a period of rapid economic and industrial change – for example, by impose controls on the industry and influencing its future direction in strategic areas of policy such as institutional investment.57
Lönnborg illustrates in Internationaliseringen av svenska försäkringsbolag (The
internationalisation of Swedish insurance companies) from 1999, how the institutional change and the need to spread risks contributed to an international entry of Swedish insurance
companies. Pearson and Lönnborg furthermore combine research on regulation and
internationalisation and investigate the impact of regulatory regimes in the internationalisation of insurance before 1914.58 Parente and Prescott argue that the lack of monopoly rights in England contributed to the economic growth in the eighteenth century.59 Along with this reasoning Pearson and Lönnborg argue the internationalisation and growth of insurance was facilitated by deregulation. Besides deregulation, internalisation also got promoted by knowledge transfer between actuaries – often trained mathematics or astrologists that had acquired their knowledge abroad and other insurance actors.60 Lindmark et. al. also stress the role of the Swedish actuarial profession in affecting the regulatory and legal infrastructure of the industry. 61
As clearly displayed, previous economic history research on the Swedish insurance industry has shown a great interest in the regulations of the insurance industry. There also seems to be consensus regarding what kind of regulations that have been of key importance for the industry, namely the insurance regulations explicit dealing with regulations of the actions of the industry. The motive, the impact and the process towards these regulations are of course of outmost importance for our understanding of the development of the Swedish insurance market. Still other regulations that have received less or no attention from scholars must also be included in the analysis, simply because there are regulations that are likely to have shaped the industry in a way that so far have passed almost unnoticed in the historical documentation of the insurance industry.
57 Lindmark, Andersson & Adams (2006), p. 351.
58 Pearson & Lönnborg (2008), pp. 59-86.
59 Parente & Prescott (2002), p. 134.
60 The first executive of Skandia had been an agent for several English insurers according to Hägg (1998) p. 150.
The corporate model of Thule was derived from a French life insurance company – Impériale. The Swedish company Victoria was influenced by the German much more successful company with the same name.
Americans also went to Britain to learn banking and finance according to Sylla (2002) pp. 287-288. According to Larsson & Lönnborg (2008, forthcoming), p. 1.: “The Swedish case demonstrates an early integration within the international insurance market and the establishment of modern domestic organisations were imitated from European standards, especially from Great Britain and Germany”.
61 Lindmark , Andersson & Adams (2006), pp. 341-370.
The lack of regulation of insurance contracting and the potential problems associated with limited property rights for married women in receiving the sum insured, attracted attention among contemporary insurance actors as early as 1883 and procedures towards an insurance contracting law were initiated. But despite this phenomenon being corresponding with the much more investigated efforts to formulate an insurance regulation, research on the regulation of insurance contracting and the connection and impact of property rights on the life insurance transaction has so far been an unexplored issue. Insurance contracting is on the other hand considered as central for the regulation of the insurance industry in the United States during the 19th century.62
Although the regulation of for example 1903 could be argued to contribute to more effective property rights by e.g. solvency restrictions and independent regulation by the State Insurance Inspectorate, the ongoing debates and problems addressed by insurance actors on the topic of insurance contracting during the 19th century, cannot be ignored. Nor the united efforts of insurance actors and the women’s movements in the ambition to strengthen the property rights for married women.
Institutional theory stresses the importance of certain regulations for a well-functioning capitalist system. The most important regulation is however according to scholars like Douglass C. North clearly identifiable property rights. Despite this, insurance researchers have not this far been concerned with comprehensive studies on how the conditions of property rights got shaped by, or shaped the insurance industry. This is especially relevant concerning one of the most fundamental issues regarding the insurance transaction – who has the right to the sum insured? It has been taken for granted that property rights in Sweden was clearly enough defined in the end of 19th century to support a financial expansion, and
therefore no further study of the impact of property rights has been conducted. But on the contrary, a large group of women, the married women to be precise, did not have sufficient property rights for the life insurance transaction or the financial system to function perfectly.
These women were certainly the intended beneficiaries when life insurance was marketed.
The ways exercised by insurance actors to propagate for improved property rights, in some cases implied membership in the women’s movement. The period of study was thus characterized both by transformation concerning financial and social matters. On the topic of social transformation, this period experienced an enhanced activity of social movements working for improved rights for blue-collar workers and women.
62 Keller (1961), p. 318.
Regarding the financial transformation, banks and insurance companies entered the market in large numbers and a financial deepening took place. Although these
transformations, the social and financial, co-existed and although insurance actors were part of these progressive, social movements, no attempt to study this interaction and to evaluate its importance has been put forward. The ambition of this licentiate thesis is therefore to suggest a theoretical framework that makes it possible to connect social and financial history by addressing the importance of useful knowledge in the development of the financial market.
This does not only imply acknowledging the general dynamics of social change in the development of the life insurance industry. It even means recognizing the importance of social movements, seemingly far from the activities of the financial market, in the process towards the take-off and maturity of the life insurance industry.
3. Useful knowledge and the financial economy
Technological innovation is according to a majority of scholars a pre-condition for economic growth. Pearson observes that compared to the attention denoted to industrial innovation, there has been limited investigation of the process of innovation in financial services.63
A point of departure of this licentiate thesis is however that social and financial history cannot be separated; and in the ambition to investigate the impact of demand side factors and emphasise the significance of social forces for financial development, this licentiate thesis will rather focus on the conditions for innovation, which according to Mokyr is useful knowledge. Mokyr further argues that the industrial revolution was not driven only by the surfacing of technological ideas, but also by the improved access to these ideas in society at large.64 The framework of this thesis is therefore to a large extent derived from the theories of Mokyr on the evolution of the knowledge economy.65
What separated earlier clusters of economic growth with the industrial revolution was according to Mokyr, theoretically, a transformation in the relationship between propositional knowledge (what), and prescriptive knowledge (how), a process which could lead to
innovations. Instead of the elite controlling the knowledge, it got offered to a general public a development that enhanced the probability and quantity of innovations, while it reduced
63 Pearson (1997), pp. 235-236.
64 Mokyr (2005), p. 285.
65 Others have stressed the relation between knowledge, technological progress and economic growth. Easterlin (1981), pp. 1-19, argues that mass schooling facilitated the acquisition and application of knowledge about new production techniques associated with the emergence of modern economic growth. Musgrave (1967), pp. 252- 253, shows how technological innovations made demands on the educational system and the labour force.
access cost, broadened the epistemic base and improved efficiency by reducing the risk of
“reinventing the wheel”.66 With the society possessing a broad epistemic base, it was a greater chance that propositional knowledge would lead to sustained improvements in prescriptive knowledge. And most importantly, these improvements did not only concern economic dimensions, it also included social improvement, alleviating the conditions for the poor and unfortunate.67
The forces behind this improved access to knowledge were according to Mokyr the impact of enlightenment.68 This in turn fostered open science and democratic ideas which added fuel to the industrial revolution.
An important concept of this study is useful knowledge. As this concept has denoted a variety of things in the economic history literature, clarification regarding its usage in this licentiate thesis is necessary.69 This thesis is going to define useful knowledge in accordance with the definition made by Mokyr. Mokyr distinguish on one hand knowledge accumulated in observing natural phenomena; by establish regularities in the environment. This definition also includes “technologies” like life insurance marketing.
The other definition concerns knowledge achieved regarding social facts and
phenomena. Mokyr claims economic knowledge (price, rates etc.) should be included in this definition, as it is necessary for efficient production and distribution.70 The argumentation of Mokyr has therefore also an obvious application on the financial expansion, of Sylla referred to as the financial revolution.
But economic knowledge is not only important for production and distribution. This study argues that useful knowledge furthermore constituted the base for the demand and consumption of life insurance. Hägg explains this situation with the prime move dilemma, which refers to a situation where consumers, because of incomplete information, are unable to rank the quality of insurers and refrain from exchange because of a too low level of trust. The prime move problem therefore may prevent the market from evolving. With other words, for a financial service or innovation to have a continuously bearing in society, the major public must have gained enough knowledge of financial matters to be able to demand e.g. a life insurance.
66 Mokyr (2002), p. 9.
67 Mokyr (2005), pp. 290-291.
68 Mokyr (2005), p. 296, defines enlightenment as: “[…] a belief in the possibility and desirability of economic progress and growth through knowledge.”
69 Among others Kuznets (1965) uses useful knowledge as synonymous with “tested” knowledge and Prescott et.
al. (2000) p. 133, talks about useful best-practice knowledge.
70 Mokyr (2002), pp. 2-4.
Sandberg (1979) also stresses in his case of “the impoverished sophisticate”, the importance of a broad epistemic base, or: “[…] a strikingly large stock of human and
institutional capital” in explaining the high GNP in mid-nineteenth century Sweden.71 He also claims that the very high elasticity of supply of financial services was directly related to the existing high levels of human capital.72
This framework therefore suggests that the key for explaining the development of the insurance industry lied in what people knew or did not know. To illustrate: An individual do not need to achieve a specific knowledge to buy e.g. shoes, that is regardless of if the shoes have been made by using the latest technology available or not. The fact is that it is still just shoes, irrespective of how they were made. Buying a life insurance demanded and still
demands of the consumer to achieve a greater level of knowledge of the service than buying a pair of shoes.73 For buying a life insurance the customer needed to know rates, premium levels and how the different kinds of insurances functioned. For buying a pair of shoes the only thing the customer needed to know was shoe size. The question is what impact this acquired knowledge for buying a life insurance had on the development of the life insurance industry in the 19th century Sweden.
In the following articles the impact and importance of useful knowledge for the
development of the life insurance industry is elaborated. The framework of useful knowledge is in the first article going to be used to investigate dynamics not only possible to refer to by using economic explanations. The dimensions of reduced access costs and the impact of a general financial knowledge in society at large will be discussed. In the second article, the framework is used to investigate how useful knowledge was diffused on a micro-level. The article focus on how useful knowledge started to entail, among other things, thoughts of the enlightenment that favoured extended rights for women and blue-collar workers. The article treats how useful knowledge gave rise to the inclusion of marginalised groups in the making of the new financial economy. For example the diffusion of the life insurance to broader groups in society and the struggle for married women’s property rights.
71 Sandberg (1979), p. 225.
72 Sandberg (1979), pp. 227-228.
73 Stalson (1942), p. 24, argues in a similar way when he discusses the importance of understanding the uncertainty of life: “We do not have to understand very much to respond to the desire for food. But we must understand to respond for, let us say credit insurance. In the former case an understanding of what is wanted is so universal that it goes unnoticed, but credit insurance could never be desired except by someone who had a wide background of training which would make him perceive the need for and the desirability of buying it”.
Article 1: Insurance and income growth: the case of
Sweden 1830-1950
Insurance and income growth: the case of Sweden 1830- 1950∗
Lars Fredrik Anderssona, Liselotte Erikssona Magnus Lindmarka,
a Department of Economic History, Umeå University, Sweden
∗ An earlier version of this paper was presented at the Swedish Economic History Conference in Stockholm, October 2007. Paper has been submitted to the Scandinavian Economic History Review.
Abstract
In this paper we investigate supply and demand factors that have been put forward to explain the growth of insurance markets during the financial revolution. We conclude that income growth, urbanisation, changes in risk and price are not sufficient to explain Swedish insurance market growth prior to 1900. Instead, we argue that cultural factors, i.e. the dismantling of ideas and business cultures based on the estates of the realms was essential for explaining the growth of insurance markets foremost between 1855 and 1900.
Key Words
Insurance history; insurance; the financial revolution; Swedish economic history; household budget survey; historical national accounts
