Managing policy lapse risk in Sweden’s life insurance market between 1915 and 1947

Abstract

We examine the challenges that Swedish life insurers faced in managing the lapse risk of policies written on the lives of the industrial urban working class between 1915 and 1947. We observe that with the threat of State socialisation of insurance in the 1930s, industrial life insurers modified their business practices to better control policy lapses. Using firm-level data, we also analyse the effect of socio-economic changes, such as rising real wages, interest rate fluctuations and unemployment on life insurance policy lapses. Our results support contemporary tests of the emergency fund and interest rate explanations for the voluntary premature termination of life insurance policies.

Keywords:

• Policy lapse risk
• life insurance
• Sweden

Notes

1. Life insurance was relatively less important in rural communities than in urban areas because farming provided supplementary sources of income and subsistence. Additionally, the support structures of agricultural communities provided a degree of cooperative-based social protection (‘social solidarity’) for rural dwellers in the event of life cycle risks, such as severe illness and/or disability.

2. The second most popular industrial life insurance in Sweden during the analysis period was fixed-term coverage, which could again be paid out at a specific time or to appointed beneficiaries. The third most popular was children’s insurance, followed by whole-of-life insurance. These policies provided payments to beneficiaries when the insured died or became infirm, and so they acted as a form of life-cycle savings that provided financial relief in times of unexpected mortality or disability. Industrial life insurance in Anglo-Saxon countries protected the insured against the stigma of a ‘pauper’s burial’. However, in Sweden, the necessity to insure against a ‘pauper’s burial’ was less urgent, since a decent funeral was less costly as the State Lutheran Church provided subsidised burial costs (Eriksson, 2010).

3. Our use of real interest rates is consistent with the interest rate hypothesis, but also their inclusion in the analysis is more realistic in the context of the present study for two main reason. First, public involvement in the Swedish stock market during the first-half of the twentieth century was very limited; alternative savings in bank deposits and bonds were far more common (Ögren, 2006). Second, the 1903 Insurance Act significantly restricted insurance company investment in equities, and so insurers’ investment portfolios comprised largely precautionary assets and fixed return securities such as government bonds. Hence, insurers’ expected future cash flows on in-force business (the so-called ‘embedded value’) and policy lapse rates were relatively more sensitive to changes in market interest rates than stock price movements.