The Feminization of a Fraternal Institution: Swedish Sickness Funds 1890-1955

The Feminization of a Fraternal Institution: Swedish Sickness Funds 1890-1955

The feminization of a fraternal institution: Swedish sickness funds 1890-1955

Liselotte Eriksson

Umeå Centre for Gender Studies

Umeå University

Lars Fredrik Andersson

Geography and Economic History

Umeå University


Since the late 19th century, Western countries have implemented a wide range of public insurance programs. In many countries, the evolution of public insurance has been modeled on cooperative, mutual insurance funds. Especially sickness insurance funds have been recognized for their importance, not only as forerunners to public insurance but also for risk sharing, risk transferring and loss mitigation. By reducing risks related to sickness, illness and accident, sickness funds improved social welfare.

In a society with a lack of sufficient medical care for the majority and where epidemics flourished due to insanitary living conditions and illnesses like pneumonia and tuberculosis caused thousands of lives annually, sickness- and burial funds performed an important task. However, a state that could imply great danger and cause both death and misery in mid nineteenth century was pregnancy. The risk to die while giving birth was 30 per cent in 1890 and an infant had a 60 per cent chance to survive its first birthday. To counteract infant mortality, the government introduced a labour protection law in 1900 that prohibited women in the industry to work until four weeks had passed after giving birth; the period was increased to six weeks in 1912. In many families there were many women that were sole providers for the family and in several working class households the wife’s salary was essential to ensure the necessary minimum income. The lack of possibility to work and provide for her family naturally affected unmarried women the most. A discourse, dominating political decisions, where unmarried mothers had no place in the public sphere and the ideology of the male breadwinner model, caused a discrepancy between reality where unmarried mothers counted for a large share and where both unmarried and married women had to work to secure a sufficient income.

Although researchers have investigated the evolution of sickness-, accident- and pension insurance few have recognized the importance of the implementation of public maternity allowance. It did e.g. not, in the case of America, see to the needs of women for maternity allowance.


Sickness insurance, feminization


250 years ago the average Swede could look forward to less than half of the life expectancy we enjoy today. The figures are primarily due to the extremely high death rate among small children and new-born’s but also to the adult population to whom infections and epidemics was a constant threat. [1] The industrialized society gave rise to increased urbanization and reliance on salaried work. At the same time the lives of industrial workers became more exposed to the risk of sickness and death than previous generations that had made their living in the country side. Unsanitary conditions and crowded quarters gave rise to epidemics and a wide spread of tuberculosis which could result in sickness and death. To meet these challenges of the average nineteenth century wage earner, sickness and burial funds were established rapidly all over Europe throughout the century.

Sickness funds have been recognized for its efficiency in controlling information asymmetry controlling such as adverse selection and moral hazard (Emery, 1996). Indeed, tests for asymmetric information in cooperative insurance societies examine how their pricing policies affected the profile of members (Gottlieb, 2007). Gottlieb argues that cooperative societies (unlike their contemporary substitutes) were able to overcome the asymmetry of information and benefit from informational advantages. The small societies and the selection mechanisms made it possible for funds to only engage in relatively known risks, reduce adverse selection; and the social control reduced fraud and moral hazard. Emery (2010) stresses that the reason for the American resistance to compulsory health insurance was that it did not offer a better alternative than voluntary alternatives. Previous research on cooperative insurance has hence focused on efficiency in sickness funds by measuring information asymmetries, aspects of equity and equality has hence been neglected. We believe this is an omission given the importance of access to insurance when discussing the social welfare implications of sickness funds.

Previous research has focused on salaried male workers – the group most likely to benefit from sickness insurance through membership in sickness funds. However, as many sickness and burial funds were institutions explicitly organized around the principle of masculine identity, the selection strategies employed by sickness- and burial funds did involve an exclusion of women from access to sickness insurance coverage (Clawson, 1983). Women were the ones most frequently excluded from sickness insurance and at the same time also among the most exposed to the risk of the loss of income due to inability to work. Women were exposed to a gender-specific health risks through delivery and childbearing. This was combined with a law were female industrial workers at the end of 1870s, in many western countries became, forbidden by law to work two or up to six weeks after giving birth, without receiving any kind of compensation. In Sweden, female industrial workers did not receive compensation until 1931 (SOU 1954:4, p. 14). If a woman was unmarried and the only breadwinner, her and her child’s situation could result in starvation and severe poverty.[2]

In this paper, we address how the institutional structure impacted on female membership in Swedish sickness funds during the period 1890-1955 – a period where the female members went from marginalization to a substantial and integrated part of the sickness insurance system. By addressing the issue of female membership, we believe that the paper may help to provide a wider understanding of the equity and equality issues inherit in social insurance. As the sickness insurance system developed from fairly small and independent and privately funded units into a centralized and subsidized insurance system under public control, the period of study offers a possibility to analyze how variation in institutional structures affected the process of feminization.

The following of the paper is organized as follows. Section two outlines the historical development of sickness funds; changes in regulatory environment and financing. Section three provides a qualitative account of the mechanisms of exclusion of women of sickness funds. Section four introduces our quantitative analysis of exclusion mechanism, while section five provides the results. Section six concludes.

Historical overview of sickness funds

The first Swedish sick and funeral insurance funds were established in the early 1700s within various trade guilds. These funds provided insurance for medical expenses and lost wages in the occurrence of illness or accident as well as for funeral costs in the event of death. In this respect, these funds are the equivalent of today’s sickness insurance providers. During the period 1870-1885, there was an increase in establishment associated with this period’s rapid industrialization and urbanization. A parliamentary committee set up in 1884 to study the emerging insurance market observed that the health coverage offered by the existing insurance funds was lacking in terms of that insurance societies frequently dissolved leaving members without insurance.

In order to encourage the development of sickness and burial funds, the 1891 Law of Registered Sick Insurance Societies implemented an administrative subsidy for those funds that voluntarily registered with the government. Registration implied sending annual financial statements as well as information concerning policy statutes to the Department of Public Administration and after 1901, to the National Board of Trade. Registered funds received a small subsidy which benefitted funds with a fewer number of members, while the subsidy per member decreased with the number of members. Hence, the regulation stimulated the development of especially small sickness funds and sickness funds showed continuous problems with reaching a large part of the population and provide safe insurance alternatives.

Public financial support (subsidies) was provided not only for administrative costs, but also for reducing membership fees and attracts new members; this in turn eventually threatened the soundness of the business (Lindeberg, 1949, p. 250). Fierce competition for members often led to detrimentally low premiums, a high rate of closures, and uncertain coverage (Berge, 1995; Edebalk, 1996). The subsequent 1910 Law of Registered Sick Insurance Societies increased regulation while maintaining the voluntary nature of the Swedish health insurance system.

In 1910 the authorities demanded a more actuarially sound relation between fees and benefits. Ex post assessments were forbidden except for temporary budget deficits and the funds were to rely primarily on regular fees. Multiple memberships in funds were banned and the burial benefit became much reduced by the law of 1910 and made many funds separate their sickness- and burial business from each other. Exclusion of members was not accepted if it was not regulated in the policy statues. The 1910’s act also demanded that a fund contended at least 100 members to get registered. The pressure on organizational changes and administrative requirements started a process of consolidation where funds merged. The increased capital requirements further led problem with exit and recruitment of members in small sickness funds. The public subsidy increased with the 1910’s act. However, the subsidy was reduced with half if a member had multiple memberships in sickness funds – a common feature in small local sickness funds (Nekby, 2004).

From 1910, sickness funds that offered maternity allowance were given subsidies. The rationale behind subsidizing funds with maternity allowance was the labor protection law of 1900 that prohibited women in industrial occupations to work six weeks after receiving a child. The labor protection law restricting the industrial work of mothers was an international convention which was put forward at the first International Labor Protection Congress in Berlin 1890 and was passed as an international convention in Washington 1919. Against the recommendations of the convention and in contrast to several other countries, Swedish women did not receive any compensation for this compulsory leave from work (Nordgren, 1933, p. 4-5). Several other countries with labor protection laws regulating the work of women, recognized the importance to compensate women in industrial work for their loss of income and introduced state funded compensations (SOU 1929:59). In Sweden however, instead of granting women compensation for their loss of income, Swedish politicians choose to give subsidies to sickness funds that offered maternity allowance as a way to encourage sickness funds to take on maternity allowance as business and stimulate the possibility for women/families affected by the labor protection law to make the necessary precautions in the event they planned to have children. A state funded compensation to women affected by the labor protection law was not realized until 1931 (SOU 1954:4, p. 14). The budget for subsidies to sickness funds constantly had to be raised due to increased demand for maternity allowance and the many sickness funds that came to offer maternity allowance.

A second major institutional change was introduced in the 1930s. In 1931, after six years of enquiry, it was settled that all sickness funds should be divided into local- and central sickness funds. Only one central sickness fund should exist in each municipality, the local funds had responsibility to collect fees from members and controlled the reported sick members. The local sickness funds had responsibility for the sickness benefits the first 21 days, after that the central sickness funds took over. A major improvement for women regarding the regulation of 1931 was that all registered sickness funds now became obliged to offer maternity allowance.

In Britain, the National Insurance Act of 1911, granted women affected by the labor protection legislation compensation, this is probably the reason to why maternity allowance in friendly societies was very rare in Britain (Brown and Taylor, 1933, p. 7). In Sweden however, where no state compensation was granted women affected by the labor protection legislation and where maternity allowance in sickness funds became state subsidized, women’s membership in sickness funds rose rapidly during the period 1900-1954, see figure 1.

The subsidies for sickness funds providing maternity allowance and later regulatory changes towards centralization and subsidizing may have contributed to the feminization of sickness funds. To examine the decline in female exclusion and rise in women’s membership in sickness funds the following section outline the data and the methods for examine how changes institutional structure impacted on the female membership in Swedish sickness funds during the period 1890-1955.

Figur 1. Female participation rate (per cent) in Swedish sickness funds 1900-1954.

Source: Registrerade sjukkassors verksamhet 1901-1910, Registrerade sjukkassor 1910-1930; Erkända Sjukkassor 1938-1954.

Examining women’s membership in sickness funds

To trace the mechanism of female exclusion and the impact of institutional changes on female participation in sickens funds we have utilized archival and official statistical data. The qualitative data comprises of archival material derived from Stockholms stadsarkiv (Stockholm City archive), Riksarkivet (Swedish National archive) and Göteborgs Landsarkiv (Goteborg Regional archive). The focus on the two first archives situated in Stockholm is partly due to that the data set used is on sickness funds in Stockholm and the archives together with the one in Goteborg hold a large share of the material available through the National Archival Database (NAD) on female sickness funds. Women’s sickness funds available and possible to identify as female sickness funds through NAD are only 23. This is a very small share of the female sickness funds evolving in the late 19th and early 20th centuries. Besides the archives searchable through NAD, there are several non-searchable, local archives holding an unknown amount of material of female sickness funds.

The official statistical data include information on each fund with respect to income, loss and balance accounts. The data give a detail account on source of income (premiums/fees, public aid and employers aid) and on expenditures (claims experiences, administrative costs, capital costs etc.). The data include information on geographical location, occupation, number of sickness days etc. From the census of sickens fund, we have employed a sample of sickness funds registered in Stockholm in our econometric analysis.

The econometric analysis is sought to empirically examine the impact of institutional structure on female participation. The basic proposition to examine is how key components of the instructional structure that change due of legislative initiative, regulatory changes and subsidies impact on female participation. Table 1 outlines the key components in the institutional structure during the period of study.

Table 1. Key components of sickness fund’s instructional structure 1900-1955

Period 1. / Period 2. / Period 3.
1892-1909 / 1910-1930 / 1931-1954
Public subsidies / X / X / X
Size of fund / X / X / X
Occupation based / X / X / X
Maternity allowance / X / X

Period 1 is governed by the 1891 act of sickness funds; Period 2 is governed by the 1910 act and period 3 the 1931 act. The first period was characterized by limited public subsidies and small sickness funds. The occupational based funds were common and frequently based on a close mutual identity of fraternalism which is expected to have excluded women. After the 1910’s act subsides and size of funds increased and also maternity allowance was provided from the mid-1910s. A more centralized structure of highly subsidised sickness funds were attained after the 1931 act.

To account for how the key components of the institutional structure impact on the female participation rate, we have measured the components as follows. Public subsidies are measured as the share of public financial support of all incomes. The size of insurance funds is measured as number of members. In addition to the size, the structure of insurance is considered during period 3 when funds was divided into centralized (national) and local insurance funds. The extent of centralization is measured as number of local funds divided by number of members. To examine if occupation-based funds excluded women, we have constructed a dummy-variable for occupation-based funds. To account for the impact of maternity allowance, we have constructed a maternity allowance variable that measures the share of maternity as share of total expenses.

Table 2. Variables and definition
Variable / Definition
Female / Female participation measured as the share of women of all members
Subsidies / Public subsidies measured as public financial support as share of total income
Size / Size of fund measured as the number of members
Decentralization / Centralized structure measured as number of local sickness funds divided by number of members
Organization / Occupational based, indicated by a dummy=1 if occupational based, 0 otherwise
Maternity / Maternity allowance measured as share of expenses on maternity as share of total expenditures

To empirically examine how the institutional components impact on the female participation rate we have employed a panel-data design as our primary strategy. As the panel and variable structure only is coherent for sub-periods, the analysis is divided for the periods given in table 1. For period 2 a further division is needed as the data on size of maternity allowance is only given for the period 1921-1931. The period 1912-1920 also suffer from omitted values on female participation and on maternity allowance. To control for the potential sample bias, a sample selection model is employed (period 2:b).

The impact of institutional structures on female participation

Table 3 report the impact of institutional components on female participation in sickness funds. During period 1 both the fixed and random effect gives the same outcome and a reasonable fit respectively. Public subsidies give a positive and significant effect on the female participation rate. We find also an expected positive impact of size and a negative impact on organization, suggesting that small and occupation oriented sickness funds exclude women to a greater extent.