Abstract

This paper deals with differences between European countries in the use of elderly care.

The objective is to find out what the consequences could be for the use of paid and unpaid care of two policy measures. The first one is a movement from a system in which the government is responsible for care towards a system in which more responsibility lies with the family. The second one is the reduction in the supply of informal care in all European countries due to the ageing of the population, the lessening of social ties and the envisioned higher employment rates of women.

The analysis is based on SHARE-data: the Survey of Health, Ageing and Retirement in Europe: a sample of 23,000 persons aged 50 years and older. We make a distinction between the use of paid care and unpaid care and acknowledge the interaction between them. Therefore we estimate a simultaneous 2-equation model in which the use of paid and unpaid care are the endogenous variables. In the first step we estimate reduced form equations of both variables simultaneously and in the second step structural equations have been estimated. The exogenous variables in our model are needs, available resources and socio-economic factors. For needs of care we constructed measures for physical, psychological and cognitive impairments. Also some impairments such as having had a stroke are included and also some subjective measures are included. For available resources we use the availability of an informal network inside and outside the own household. Socio-economic factors are instrumented by level of education, income, age, gender and urbanisation grade.

We find that the influence of the use of unpaid care on the use of paid care and vice versa differ between the various European regions. In Northern Europe the use of unpaid care is complementary to the use of paid care, in Southern Europe the use of paid care is more a substitute to the use of unpaid care. We have applied BlinderOaxaca-related methods to decompose the differences between European regions into a part attributable to observable characteristics and a part attributable to behavioural effects. We show that around 10% of the 15 percentage points difference in the use of paid care between the Northern and Southern European countries is explained by differences in observables as demography, socioeconomics and health. But 90 per cent is due to differences in responses to the observables, which we interpret as differences in the health care systems and individual preferences. The difference in the use of unpaid care between the Northern and Southern European countries is 8 percentage points, where based on differences in observable characteristics a negative difference was expected.

So the difference in the use of unpaid care is totally due to a difference in behavioural effects with respect to differences in care systems. Simulations show that moving towards a more Southern European health care system would result in a significant reduction of in the use of both paid and unpaid care in Northern Europe.

Slides