Academic journal article Journal of Risk and Insurance

Pure versus Mutual Health Insurance: Evidence from Swedish Historical Data

Academic journal article Journal of Risk and Insurance

Pure versus Mutual Health Insurance: Evidence from Swedish Historical Data

Article excerpt


Using data from voluntary Swedish health insurance societies 1902-1910, this article analyzes the coexistence of pure and mutual insurance societies where pure societies are characterized by charging ex ante premiums only while mutuals in addition charge ex post assessments. On average, mutual insurance societies are found to be larger and to offer greater insurance coverage per member. Pure insurance societies have, on average, higher insurance coverage per day, greater mean levels of moral hazard controls, a higher mean number of policy categories, and greater longevity.


This study uses a unique historic data set to analyze the coexistence of pure and mutual insurance societies. Although there is a fairly extensive literature on the theoretical differences between these insurance types as well as the nature of their coexistence (see Born et al., 1995; Marshall, 1974; Skogh, 1999; Smith and Stultzer, 1995; van der Linden, 1996; Wu, 2000), empirical work is scant. Pure and mutual insurance are defined by their choice of financing system. Pure insurance societies use ex ante premiums only whereas mutual insurance societies in addition to ex ante premiums, also excise assessments ex post. (1) Mutual insurance is therefore a form of risk sharing between members. This type of insurance is expected to cater to smaller groups of homogenous members, in terms of risk, motivated to control moral hazard problems. The fixed contracts offered by pure insurance societies cater, in turn, to a more heterogenous, potentially higher-risk clientele which without specific controls, lacks the motivation to minimize moral hazard problems. The purpose of this study is to empirically examine theoretical predictions concerning pure and mutual insurance, as defined by a type of financing system employed.

The data, covering all Swedish sick and funeral insurance societies (2) from 1892 to 1910, were compiled as a result of the 1891 Law of Registered Sick Insurance Societies offering an administration subsidy to all insurance societies that freely registered with the government. As such, detailed information concerning both financial and policy variables of insurance societies became available. The pre-1910 insurance market in Sweden was characterized by free entry and exit and was largely unregulated. Insurance societies were free to set their own insurance contracts, premium, and assessment levels, and to determine the type of financial system. This data set therefore provides a unique opportunity to examine insurance theory within the framework of an unregulated and competitive insurance market. The results can shed light on issues concerning the impact of regulation/deregulation on the insurance market by providing one study of the deregulated case. Regulation may otherwise influence the underlying conditions of the insurance market to sustain both insurance types.

The insurance society data provide indicators for variables of interest that are used to analyze differences between pure and mutual insurance. To test the robustness of the initial univariate results when controlling for society membership size and differences in registered sick levels (among others), a logit analysis is performed estimating the probability of insurance society i being pure. Finally, in order to check that systematic differences between societies with longer or shorter registration periods do not influence the results, an analysis is run on a balanced subpanel of insurance societies.

The main results are that mutual insurance societies are not constrained by smaller membership sizes. Pure insurance societies provide greater coverage per day whereas mutuals are, on average, associated with more generous insurance policies per member. Consistent with theoretical predictions, pure insurance societies are found to have more stringent moral hazard controls, a greater mean number of policy categories, and are, on average, olden

The remainder of this article is as follows. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.