Who is willing to pay for long-term care insurance in Catalonia?

Abstract

Both public and private insurance for long-term care (LTC) in European countries employed is underdeveloped. However, limited evidence is gathered on the behavioural demand responses to the introduction of insurance schemes expanding insurance coverage. This paper aims at empirically examining the determinants of the ex ante demand for LTC coverage drawing upon an insurance-based WTP referendum approach format. Data is collected from a representative population sample of Catalonia (Spain), before a countrywide funding system was developed. Our findings suggest that (based on our questionnaire format) only one-fifth of the population is willing to pay for LTC coverage though it is a highly elastic product. Ex ante demand for LTC coverage is driven by individual's perceptions of old age disability risk (private information asymmetry) and housing tenure (giving rise to self-insurance), the latter reduces the probability of insurance coverage demand.

Introduction

The financing of long-term care (LTC) for older people is a top health policy concern in Europe with population ageing. European Commission supported research [1] concludes that with few exceptions, there is inadequate insurance coverage against the risks of long-term care in some European Union countries. This lies in the fact that costs of LTC can be individually ‘catastrophic’ when severe dependency requires long-term assistance for personal and domestic care not provided by family members. In some countries such as Spain, the size of the dependant old age population is expected to double over the next 50 years [15]. This feature results from changes in family structures (e.g., smaller family size and the full women in the labour market) pointing towards an ostensible reduction in the availability of informal caregivers and anticipating a higher demand for formal care. Although public financing tends to be means tested and subject to sizeable co-payments, the cost of paying LTC out-of-pocket in Spain stands as a major concern for many middle-income couples that already foresee that might not qualify for full state help.

Though the potential demand for LTC coverage (both public and/or private), Spain has only recently defined the lines of what is supposed to be a so-called National Long-Term Care System starting in 2007. That is a tax-based system managed at the regional level though not fully operative up until 2015. However, still public coverage remains means tested, which gives rise to the development of some forms of private financing, including private insurance. The possibility of private financing alternatives are limited evidence from the insurance market in the United States (US) suggest that only around 10–20% of the elderly can afford private long-term care insurance [2], [3], [4]¹ which is interpreted as evidence of a slow market development [5], [6]. Some explanations refer to the reluctance among middle age individuals to purchase LTC insurance [7] and the existence of asymmetric information [8]. In a seminal paper, Pauly [9] suggests that Medicaid crowds-out LTC insurance by creating incentives to exhaust one's individual savings to be eligible for public funding. Other explanations for the limited market development refer to the underlying effects on intergenerational decisions such as bequests [10] along with intra-family moral hazard [6]. These arguments have been empirical and striking evidence find that unobserved information, and most notably individual's risk perceptions, exhibit an opposite correlation with insurance coverage [11]. This finding is consistent with previous studies [12] indicating that risk underestimation of nursing home entry does influence the purchase of LTC insurance as well as other evidence indicating that education and knowledge are the primary factors explaining LTC insurance among middle-income subjects [13].

Taken previous empirical evidence together, we contribute to the literature by empirically estimating the willingness to pay for covering LTC in Spain, in a time where no insurance entitlement existed. We estimate the share of the population that would be willing to purchase hypothetical LTC insurance policy at different premium bids.² Under the assumption of no biased responses, this provides an indication of hypothetical ex ante demand function for LTC insurance coverage.

Second, we examine the influence of risk perceptions of disability in old age. If risk perceptions, which are unobserved preference-related characteristics, are found to explain purchase decisions, this would provide suggestive evidence in line with previous evidence [11]. The importance examining effect of the disability risk underestimation in an ex ante LTC insurance demand format lies in that it strengthens the hypothesis of individuals’ early ‘myopia’ in anticipating LTC needs [5]. Third, on the basis of our exercise we examine other important determinants such as the role of private finance substitutes such as housing property, which provide a source of self-insurance. Finally, we test the effect of the insurance premium—as standard demand theory would predict. Indeed, this enables to elucidate whether LTC insurance is a price-elastic good. That being the case would suggest that tax-incentives might play a role in the expansion of potential LTC insurance coverage.

Currently, Spain allocates about 0.65% of its GDP to community and institutional care and it is expected to double in the next two decades [14]. Spanish LTC relies heavily on informal care, as 75% of dependent elderly receive no formal attention [15] and access to formal social services is based on needs and means assessment, especially for residential care. Co-payments are significant, accounting for 25% of pension income for day care and 75% residential care, respectively. Public expenditure on social care is comparatively low, accounting for no more than 0.22% of GDP, a situation that reveals a considerable gap between health and social care in terms of both provision and funding in Spain.

The paper is organised as follows. In Section 2 we briefly summarize previous literature on the determinants of LTC insurance purchase. Section 3 gives an outline of the data and Section 4 reports the empirical methods. In Section 5 we present the results on the basis of the hypothesis set in the empirical model. The paper ends with some concluding remarks and policy implications.