Abstract

In developed countries, individuals having difficulties to perform the activities of daily living may benefit from public home care subsidies. Such subsidies may cover part of the cost of professional home care services or, in some cases, be used to compensate informal caregivers. One notable and distinctive feature of French long-term care public policies is that they distinguish between handicap schemes, accessible to individuals below age 60, and dependence schemes, open to those aged 60 or more. Public debate has been going on around this ‘age 60 barrier’, some voices raising the concern that it could induce unfair compensation of the expenditure costs of disability. However, handicap and dependence schemes differ in many ways – eligibility conditions, type of care that can be subsidized, copayment rates -, making it difficult to compare them. We offer to exploit individual survey data and econometric techniques to assess whether this institutional threshold has an impact on non-medical home care utilization rates among the community-dwelling disabled population.

We use the French Health and Disability Survey on Households (HSM) to get a sample of individuals aged 50 to 74 with restrictions in at least one IADL or ADL. We use a Regression Discontinuity Design (RDD), considering the age 60 as being the threshold besides which dependence schemes become accessible. We fit a bivariate probit model to account for the simultaneity of formal and informal home care utilization decisions. We also control for a large set of individual and family characteristics that could affect home care utilization independently from the difference in public schemes available.

We find that, conditional on living in the community, being a ‘dependent elderly’ rather than a ‘handicapped adult’ increases the probability to use non-medical professional home care by 6 to 10 percentage points. This effect is large given that only 7% of individuals aged 50 to 59 receive formal home care. The decrease of informal care utilization around the threshold is of smaller magnitude (2 to 4 percentage points) and less robust, but is suggestive of small substitution effects, consistent with what is found in the literature. However, the probability to receive any home care does not appear to be affected.

Using the same RDD strategy and a complementary dataset on the population living in institution, we also find evidence that the institutional threshold of age 60 affect living arrangements. The probability to live in institution rather than in the community increases by around 3 percentage points at age 60.

Overall, these results show that the institutional barrier at age 60 has substantial impacts on the way individuals’ restrictions in the activities of daily living are being compensated. This is new evidence that care arrangements do respond to long-term care public schemes. Our results also contribute to the growing literature that documents the importance of the design of long-term care policies on care utilization. Finally, our paper has important implications for French home care policies, as the right to equal disability compensation irrespective of age does not seem to be guaranteed.

Slides