1 Whichever way you look at it, pensions are a problem. The Policy Studies Institute has expressed concern at the increased burden of State pensions fifty years from now when the number of pensioners increases in relation to the number of contributors. The Institute for Fiscal Studies too is anxious for the same reason and makes the criticism that insufficient consideration was given to the long-term cost when the State scheme was extended in 1978 so as to provide an additional earnings-related pension. The report of the Scott Committee has reduced but not wholly eliminated controversy on the extra value of index-linked public service pensions compared with the pensions commonly provided in the private sector. Employers, on whom the financial burden finally rests, are hard pressed to finance the accrued pension rights of their employees in final salary pension schemes and are unable to make the additional contribution necessary to preserve the values of the pensions and deferred pensions of former employees.

1.1. In a recent paper, Pollard(1) introduced the morbidity-mortality table as a model for measuring the interaction between morbidity and mortality of chronic degenerative conditions like breast cancer and lung cancer. This device consists essentially of two double decrement tables: for example, persons either remain free of breast cancer for life or depart from the breast cancer-free group by incurring breast cancer or by dying from other causes; those who incur breast cancer either remain alive with it until they die from it or die from other causes. Clearly this model is very similar to the combined marriage-mortality table which is referred to briefly in U.K. actuarial texts(2). Neither of these models, however, allows for increments, for example recoveries from breast cancer in Pollard's morbidity-mortality table. In recent years one of the major advances in the theoretical development and practical construction of life tables has been the incorporation of increments. A guide to this methodology is provided here with particular emphasis on the measurement of incidence and prevalence of disease. Further applications are given in an accompanying paper, to appear in Volume 111(3).

This paper finds a strong size effect for the total business of 106 U.K. life companies from the 1981–82 Insurance Directory and Year Book.

The industry is regulated by the 1982 Insurance Companies Act which consolidated earlier legislation. This does not control output or prices of firms but many other items including investment. The industry is not highly concentrated with the five largest firms in 1980 with 30% of premiums. In 1980, total ordinary branch premiums amounted to £5.8 billion.

Section 2 surveys briefly some relevant research, while §3 discusses the data and size and other economic factors for life assurance firms. Section 4 discusses the empirical results while §5 has a brief conclusion.