Kinship ties become more important with age, as kin form a major resource in times of need, but there is concern that ties are weakening, and the numbers and types of kin may vary substantially from the relatively benign current position. Having kin is a precondition for kin interaction and support, and cumulative lifetime experiences such as total number of partnerships substantially affect interactions with other kin. This part of the project will produce projections of the numbers of people with particular types of kin such as spouses, living children, grandchildren and sibs, including step- and half-relatives.

Aims

The work being undertaken falls into three main parts

Providing estimates of number of kin distinguishing important categories, such as partners and numbers of living children by sex and marital status of both generations. This is necessary, since marital status is an important determinant of household structure among older people, and, for example, existence of living children is likely to have very different implications for care of elderly men who divorced when the child was small or who remarried, compared with widowed men. This involves further development of methods to analyse such kinship networks.
Development of models of shared frailty, and assessment of the sensitivity of results to such effects. This involves estimation of such effects in Britain. The sensitivity of results to assortative mating, and to intergenerational transmission of demographic behaviours is being tested.
Kinship microsimulation models have principally been used to model total populations rather than differentials between socio-economic groups, we are producing estimates of the likely different life patterns of people in various educational level categories, for example, how many years are likely to be spent in the widowed state, or without living children etc., and how they differ by gender.

Methods

We are using a proven microsimulation technology, the SOCSIM model developed by Gene Hammel and Ken Wachter at Berkeley. The model is being further enhanced to incorporate explicit feedback effects, and, in particular, to permit the characteristics of spouses and kin members to be correlated. Such shared frailty models are particularly important at older ages, since the distribution of marital status will be different if those who are likely to survive tend to have partners who themselves have higher expected longevity. Current approaches do not take account of such issues.

Linkages

The marital status of older people is a major determinant of economic conditions, especially for women, and of availability of informal co-resident care, especially for men, but projections of long-term care need have not systematically included such factors, therefore these are relevant to Work Packages 4 and 5. Projection of kinship configurations provides a basis for numerical assessment of kinship ‘resources’ and kinship ‘burdens’ which underpin patterns of co-residence, financial aid, practical assistance in the face of illness or disability, and emotional nurture, given and received in Work Package 4, as well as to the overall modelling of Work Package 5.

Outputs

As well as contributing to the other models, these results are of interest in their own right, informing debate in a number of disciplines such as demography, history, anthropology, sociology, social policy, forecasting, epidemiology and gerontology. The current version of the model is publicly available and in use: the enhanced model will be placed in the public domain at the end of the project.

Contact

Professor Mike Murphy,
Department of Social Policy,
London School of Economics,
Houghton Street,
London WC2A 2AE
UK

Tel: 020-7955-7661 (direct line)
Email: m.murphy@lse.ac.uk

^ Back to top