23 Figure 2.3: Rectangularization: Survival Curve for Men in Germany and Hypothetical Accessible Physiological Lifetime. 0 20 40 60 80 100 0 10 20 30 40 50 60 70 80 90 100Proportion of Survivors Accessible Age Accessible Physiological Lifetime 1982/84 West Germany 1949/51 West Germany 1871/80 German Reich Source: Schramm (1996), p. 25. Moreover, in contrast to the medicalisation thesis, Fries presumes a “compression of morbidity” (Fries 1980, p. 133) which leads not only to rectangularization of the mortality curve, but also of the morbidity curve. This means that medical technological progress and a healthier way of living would result in postponement of chronic diseases and improvement of people’s health status when they are younger. If the occurrence of chronic diseases is more protracted than lifespan increases, then the time of disability and senescence will become shorter in absolute terms. This is called the “absolute compression thesis”. A more moderate version is the “relative compression thesis”, which assumes that only the proportion of lifetime which is spent in chronic illness declines because of the increasing life expectancy. Graphically, with the compression thesis the expenditure pro? le would become smoother in lower ages and steeper in the highest ages (? gure 2.2). 2.3.3. Empirical Evidence The impacts of the medicalisation and the compression theses on the development of total health care expenditure are completely different. Compared to the medicalisation thesis, a status-quo outlook on the basis of present expenditure pro? les would underestimate the true growth development. The compression thesis would exactly predict the opposite and even decreasing per capita health expenditure could be possible (Breyer and Felder 2004, p. 4). As yet, there does not exist any evidence for either one of the theses: the empirical studies which have been carried out have deliver controversial results. Many surveys support the compression thesis: Getzen (1992) accomplished cross-sectional as well as time series analyses based on OECD health data from 20 countries from 1960 to 1988. In an econometric regression he tested real per capita health expenditure on age 24 structure, which is measured by the percentage of the population 65 and over, and growth rates in GNP. He came to the conclusion that ageing is not a signi? cant cause of rising health costs. This result stands in opposition to the medicalisation thesis. Moreover, Dinkel (1998) found indications that the average state of health of the German population, especially of older people, improved between 1978 and 1995. For the UK, Seshamani and Gray (2002) ascertained a decreasing proportion of National Health Service’s (NHS’s) expenditure allocated to older people. Both studies, therefore, disagree with Verbrugge’s supposition that a longer life leads to worsening health. Furthermore, con? rming the residual life-time effect, Lubitz and Riley (1993) showed that between 1976 and 1988 US Medicare costs for people in their last year of life, who accounted for an almost constant rate of around 5 % of all bene? ciaries, ? uctuated between 27.7 and 30.6 % of the total Medicare budget. Per capita expenditure for decedents was seven times higher than for survivors in the same age group.5 Similar outcomes were presented by Zweifel et al. (1999) for Switzerland. Longitudinal data from 1983 to 1992 indicate that health care expenditure depend on remaining time to death rather than calendar age.6 Comparing the number of hospital days, Busse et al. (2002) concluded that people in their last three years of life “consume” the more hospital days the younger they are. Likewise were the results of Hogan et al. (2001).7 Taken together proponents of the compression thesis try to emphasise the in? uence of mortality costs on total health care expenditure. In contrast, advocates of the medicalisation thesis refer to higher expenditure growth rates in older age-groups. Brenner (1992) stated that ageing is consistent with multimorbidity since on average people who are older than 85 years show twice as many diseases than people between 25 and 34 years. Wille and Erbsland (1993) showed that the proportion of expenditure for retirees on total SHI health disbursements increased from 26.1 % in 1970 to 41.4 % in 1991, although the quota of retired to non-retired persons remained almost constant. Diverse effects found Nocera (1996) concluding on Swiss data that outpatient cost development corresponds with the compression thesis while for inpatient treatments the medicalisation thesis appears to be con? rmed. For US Medicare spending, Cutler and Meara (1999) exhibited a real per capita expenditure growth for people above the age of 85 by 4 %, while those between 65 and 69 years rose only by 2 % annually. Likewise Fuchs (1998) stated that change rates tend to be greater at older ages. Comparable developments were detected by Buchner (2002) and Buchner and Wasem (2004) for Germany. With data from the largest German private health insurer, they show that the expenditure pro? les became steeper in the past, i.e. that per capita ex- 5 See also Lubitz et al. (1995) and Hogan et al. (2001) for similar Medicare results. 6 This study has initiated a discussion about the appropriateness of the chosen econometric methodology. Salas and Raftery (2001) and Seshamani and Gray (2004a, 2004b) challenged the robustness of the results. Felder et al. (2000), Zweifel et al. (2001) and Zweifel et al. (2004) tried to use modi? ed approaches to con? rm the original results. 7 However, empirical studies demonstrate that in practice physicians to some extent conduct agerelated rationing, i.e. less intensive treatment for older persons compared to younger ones in cases of same diseases (Brockman 2002). Such an age-related curtailment decreases expenditure for older people and could give misleading evidence for the compression thesis. 25 penditure for older people grew faster than for younger people. Relating to multimorbidity, Naegele (1999) con? rmed that there exists a typical age-related morbidity.8 This brief overview of empirical surveys demonstrates the ongoing debate about the in? uence of the ageing population on the development of health care expenditure. At the moment, only two observations seem to be established: ? rstly, a great proportion of all lifetime health care expenditure accrues in the last months before death. This speaks in favour of the compression thesis, but this insight on its own does not answer the question of how prolonged expectancy of life will in? uence total health care disbursements. Even hypothetical cost savings from delay in morbidity need not be large enough to compensate for cost increases due to potential multimorbidity in the highest age-groups (Miller 2000). Secondly, per capita health care expenditure for older people has grown faster than for younger people in the past. Wasem (1995) called this observation “steeping“ of health care expenditure pro? les. Unless concurrent expenditure for younger people decreases, this observation speaks in favour of the medicalisation thesis. 2.4. Medicalisation and Compression Theses in the Pharmaceutical Sector As long as comprehensive individual medical records are not available, the best conceivable method to ? nd indication for the dynamic relationship between age and disbursements is to analyse the development of the age-related expenditure pro? les. For the pharmaceutical market, Buchner (2002) found that the SHI drug expenditure pro- ? le remained almost constant between 1988 and 1998. This would imply that neither the medicalisation nor compression effect exists. But medications are highly heterogeneous goods which are used against a great number of different disturbances of health. It can be supposed that some drugs are mainly prescribed for diseases which normally occur in the last months of life and others are used largely independently from distance to death and there is an incrementally greater consumption at higher ages. If the trends of the various drugs cancel each other out, looking at total expenditure conceals those different developments. Even if the total expenditure pro? le seem to be unaffected by the ageing population, it could be possible that the pharmaceutical sector partly develops accordingly to both the medicalisation and the compression theses.9 Since results of expenditure forecasts differ depending on whether a 8 Strengthening of the medicalisation thesis could also follow from the so-called “Sisyphus-Syndrome”. According to that theory, medical achievements support the ageing population and could lead to changes of political majorities. If older people win a majority, more political pressure towards comprehensive health care provision could accrue. Zweifel (1989) did not found empirical evidence for this theory. Zweifel et al. (2005) suggest that a “Sisyphus-Syndrome” may have been operative in the OECD countries until the end of the past century. 9 This supposition coincides with the “bi-modal” concept (Kane 1988, SVR 2004) which constitutes a middle-way between both disputed theses. It is based on the assumption that, on average, the population’s health status continuously improves, but concurrently the number of disabled