Elsevier

Health Policy

Volume 80, Issue 1, January 2007, Pages 107-134
Health Policy

Long term care financing in four OECD countries: Fiscal burden and distributive effects

https://doi.org/10.1016/j.healthpol.2006.02.002Get rights and content

Abstract

This paper compares long term care (LTC) systems in four OECD countries (UK, Japan, Sweden and Germany). In the UK, provision is means tested, so that out of pocket payments depend on levels of income, savings and assets. In Sweden, where the system is wholly tax-financed, provision is essentially free at the point of use. In Germany and Japan, provision is financed from recently introduced compulsory insurance schemes, although the details of how each scheme operates and the distributive consequences differ somewhat. The paper analyses the effects of importing the other three countries’ systems for financing LTC into the UK, focussing on both the distributive consequences and the tax burden. It finds that the German system would not be an improvement on the current UK system, because it uses a regressive method of financing. Therefore, the discussion of possible alternatives to the present UK system could be restricted to a general tax-based system as used in Sweden or the compulsory insurance system as used in Japan. The results suggest that all three systems would imply increased taxes in the UK.

Introduction

As with public pension schemes, public systems for long term care (LTC) redistribute resources between generations, genders and people in different income groups. Accordingly, the implications of different models of financial provision need to be analysed in terms of sustainability and social justice. Do different systems fare differently as the ageing process takes off in the near future? Who are the typical gainers and losers from different systems? It is the purpose of this paper to analyse these and related issues, using a disability projection model developed by Rickayzen and Walsh [1].

There is a considerable variety in the systems developed in different developed countries to provide long term care (LTC) for the elderly. In all developed countries, citizens in need of care, who cannot afford long term care by other means, are given publicly financed care. On top of this minimum state intervention, there is a wide range in how much a state is involved in the financing and provision of long term care. The Scandinavian countries have developed models where local authorities provide services virtually free of charge, whereas the U.K. and the U.S. tend to give the state the residual role to finance LTC once other potential sources have been exhausted. Most other countries are somewhere between those two extremes (cf. [2]).

These differences in the role of the state have implications for the aggregate costs. In Sweden, total public expenditure on LTC for elderly comes to 3.2% of GDP [3]. This is more than three times as much as in the U.S., where total public expenditure is less than 1% of GDP (cf. [4]). Most countries lie somewhere in between. For instance, in the U.K. around 1% of GDP is contributed from the public purse each year [5]. The cost to the state is a function of the proportion of the population who are elderly people requiring LTC, as well as the actual cost of care.

It is unclear, though, to what extent those differences are due to different degrees of state intervention, since the national systems differ in many other ways as well. Each country has its own definition of LTC, and each country offers its particular package of services. Furthermore, relative prices of LTC tend to be different in different countries. On top of that, assessment of need and eligibility criteria-both financial and in terms of disability-differ, as well as the duties assigned by law to spouses and children.

These differences in the design of the systems clearly have important implications. The aggregate costs of publicly financed LTC, which vary between the countries right from the outset, will probably evolve in different ways as demographic changes occur. Furthermore, a public LTC system implies that there is redistribution of resources going on (the most important one being between healthy and disabled people, but also in other dimensions). Given the considerable differences between the various LTC systems, one would expect the distributive effects of the systems also to differ quite substantially.

The purpose of this paper is to deal with those two issues, i.e. cost implications and distributive effects. We want to make comparative forecasts on how total public spending would evolve if different models of LTC provision were tried out in the same country. Furthermore, we would like to determine who are typically the main ‘gainers’ and ‘losers’ from the different systems. This is done by calculating the net present value (NPV) and ’money's worth’ of the public LTC system in different countries. The former measure tends to be more interesting from a societal point of view: accounting for the amount of redistribution taking place in the system, whereas the latter is of more interest from the individual perspective.

Our study covers the LTC systems in four different countries: Japan, Germany, Sweden and the U.K. There are three main reasons for selecting these countries. Firstly, it is evident that they are all taking the costs of provision and the means of financing very seriously, since they have each undertaken a range of reforms in recent years. Ageing is proceeding rapidly in each country and the numbers of elderly are predicted to grow substantially. On the other hand, the policy responses have been very different, as well as starting from different bases and social traditions.

This paper is a contribution to an emerging literature on comparative advantages of different systems for financing LTC. Wittenberg et al. [6] provide a good theoretical overview of the policy options and the tradeoffs connected with designing a system for LTC financing and provision. Comas-Herrera et al. [7] analyse the impact of population ageing on public spending in four European countries. In addition, they analyse the implications of introducing a more comprehensive LTC system in countries where the support system has traditionally been relying on informal care and private funding. The main difference between that study and ours is that we analyse the economic implications of different systems on public spending and equity, when applied to the same country-the United Kingdom. Furthermore, our disability projection model is a multiple state model whereas the models used by Comas-Herrera et al. [7] are cell-based macrosimulation models.

Here, we present an alternative to existing methodologies for analysing the long term effects of policy reform on the LTC sector. The main advantage of our approach is that the disability projection model is based on individual transition rates from one level of disability (including “no disability”) to another. Hence, individual trajectories of disability can be followed over the entire life cycle. Using this approach, we can make a complete estimate of the contributions paid to, and the benefits received from, the public support system for different types of individuals. This information can be used to compare the net balance of different types of individuals in order to study whether any policy regime systematically benefits or disadvantages any particular group.

It should be noted that our modelling approach involves studying the LTC system in a transitory phase, during which many people have already made contributions to, and received benefits from, the ‘old’ system. Whilst it would be possible to project and compare scenarios for ‘mature’ systems, in practice it makes very little sense. This is because the transition phase to a new system could take as long as 100 years to complete, and so public policy is far more likely to be concerned with the distributive effects in terms of gainers and losers in the transitional phase, than in the very long run, and it is clear from our analysis that these are strongly system-dependent.

Existing models for projections of future needs for LTC are either cell based macrosimulation models or microsimulation models (cf. [8], [9], [10], [11]). For the kind of analysis we undertake in this paper, microsimulation would be the most reasonable alternative to our approach. Microsimulation has several well known positive features: it enables the researcher to undertake a very detailed analysis, and it can examine intertemporal issues and simulate policy changes in issues that require historical information. Most importantly, however, a microsimulation analysis would allow for behavioural responses in a way that we are not able to do [12]. On the other hand the incorporation of more detailed behavioural assumptions would seem to be of questionable value and appropriateness in relation to processes and timescales operating over long periods, as is the case in the analysis presented in this paper. Hence, microsimulation runs the risk of giving an illusion of realism that may, in fact, be unfounded.

Instead, our modelling approach is based on the assumption that the crucial factor driving demand for LTC is the underlying level of dependency (i.e. need). Accordingly, we employ a disability model which allows for a very detailed analysis of prevalence at different ages and at different points in time. We make the remaining assumptions as simple, but realistic, as possible.

Our approach is able to deliver some interesting results. We find that despite substantial differences in the parameters of the various systems, the proportional increase in the contribution rates needed to keep the systems in balance are roughly equal. During the next four decades, they increase by around 30% in all systems. This means that the contribution rate needed will exceed 3% in some cases, whereas the UK system stays at a conveniently low level throughout the projection period. The only exception to this general increase is Japan, where the financial contributions from people aged 40 and over increase by much less, around 20%. This finding highlights an important dilemma facing the policy maker. Putting a relatively higher proportion of LTC financing on older people, who increase in numbers, is a good way to keep cost increases down. However, a less comprehensive social insurance system-such as in Japan-means less mitigation of risks related to health and longevity.

It is no surprise that all systems entail some redistribution from men to women – largely due to differences in health and longevity. However, over the life cycle, the Swedish and the Japanese systems redistribute more than twice as much as the UK system. This is an important aspect of LTC financing that seems to have been overlooked.

The paper is organised as follows. In Section 2 we give a brief overview of the LTC systems in the four countries which are included in our study. In Section 3 we present our methodology and data. Section 4 presents the results and Section 5 the conclusions. A thorough description of the disability model used for projections is provided in Appendix A, a sensitivity analysis is provided in Appendix B and more detailed information on the structure of the different LTC systems is provided in Appendix C.

Section snippets

Overview of LTC systems

There is a wide variety of LTC systems at work in the developed world. Countries have generally chosen very different paths and reforms have normally borrowed inspiration more from national traditions in the realms of health care and public pensions, than from other countries’ models (cf. [13]). LTC systems may be evaluated in many dimensions and there is thus a multitude of possibilities for public policy. As suggested by Wittenberg et al. [6], the most important decisions that policy-makers

Methodology

Our main focus in this paper is to analyse the distributive impact of different systems to finance LTC. There are many possible dimensions in which this comparison could be made, and there are several methodologies available. We have decided to focus on three main dimensions, namely income, gender and age. Thus, we want to see how different systems of financing LTC redistribute between different income groups (high versus low income earners), men and women, and people of different ages.

Our

Contribution rates

First we consider the implied contribution rates plus tax rates associated with the costs and benefits of each system. The public cost of each system is expressed as the combined contribution plus tax rate that would be necessary for the system to break even year on year. A comparison of the effects is given in Fig. 2. For Japan and Germany, the general tax rate applying to all is shown as well as the higher rate applying to some groups.

Fig. 2 reveals that the UK system is by far the ‘cheapest’

Conclusion

The purpose of this paper has been to explore the effects of ‘importing’ into the UK another country's system for financing LTC, with particular emphasis on the distributive effects. This is a complex issue to analyse as LTC systems involve very long time perspectives and redistribution in various dimensions. We have chosen to focus on the prevalence of disability as the main driving force behind LTC costs, and made use of a sophisticated disability projection model that allows us to follow

References (35)

  • M. Karlsson et al.

    Future costs for long-term care. Cost projections for long-term care for older people in the United Kingdom

    Health Policy

    (2006)
  • B. Rickayzen et al.

    A multi-state model of disability for the U.K.: implications for need for long term care for the elderly

    British Actuarial Journal

    (2002)
  • Karlsson, M. Comparative analysis of long-term care systems in four countries, IIASA interim report IR-02-003. Vienna:...
  • Socialstyrelsen. Jämförelsetal för socialtjänsten 2001. Stockholm: Socialstyrelsen;...
  • J. Feder et al.

    Long-term care in the United States: an overview

    Health Affairs

    (2000)
  • Laing & Buisson. Care of elderly people. Market survey 2001. London: Laing & Buisson;...
  • R. Wittenberg et al.

    Funding long-term care: the public and private options

  • Comas-Herrera A, Costa-Font J, Gori C, Di Maio A, Patxot C, Pickard L, et al. European study of long-term care...
  • S.R. Nutall

    Financing long-term care in Great Britain

    Journal of the Institute of Actuaries

    (1994)
  • Wittenberg R, Pickard L, Comas-Herrera A, Davies B, Darton R. Future demand for long-term care, 2001 to 2031....
  • E. Richards

    Paying for long term care

    (1996)
  • R. Hancock et al.

    Who will pay for long-term care in the UK? Projections linking macro-and micro-simulation models

    Fiscal Studies

    (2003)
  • C. O’Donoghue

    Dynamic microsimulation: a methodological survey

    Brazilian Electronic Journal of Economics

    (2001)
  • X. Scheil-Adlung

    Social security for dependent persons in Germany and other countries: between tradition and innovation

    International Social Security Review

    (1995)
  • A. Evers

    The new long-term care insurance program in Germany

    Journal of Ageing & Social Policy

    (1998)
  • Bundesministerium für Gesundheit. Zahlen und Fakten zur Pflegeversicherung. Berlin: Bundesregierung;...
  • Campbell J, Ikegami N. Designing an independent LTC system: in General and in Japan. Discussion paper;...
  • Cited by (24)

    View all citing articles on Scopus
    View full text