John Vincent, University of Exeter, U.K.
This paper presents a critique of influential work by Oeppen and Vaupel (2002). Their work highlights weaknesses in projections of life expectancy and hypothesizes linear extensions of maximum life expectancy into the future. Their projection of an increase of one extra years life expectancy in every five years for the most rapidly ageing populations, has gained wide currency.
Such demographic analysis has been highly influential in changing
political and policy orientations at national and international level. In
particular, it has influenced actuaries and pension policy makers in societies
with ageing populations. In the
However, Oeppen and Vaupel’s analysis has considerable methodological and other limitations and is part of a significant controversy over the future possibilities for life extension within demography, bio-gerontology and the broader scientific community. Key points at issue include appropriate methods for projection of demographic trends, misunderstanding of the findings and interpretation of the fundamental biology of the ageing process. It is important that key policy issues are not de-politicised by use of apparently neutral demographic analysis and that the social, political and cultural assumptions behind them are revealed and understood.
Projections of future demographic trends in terms of ageing populations are crucial in the process of forming public policies, insitutional and commercial strategies. Attitudes and expectations of the future influence the ways in which current older people are viewed. Demographers can be classified into three kinds on the basis of their expectations about the future of life expectancy.
“In a simple three-way classification of the major positions that is used by the present authors (Olshansky and Carnes 2007; Olshansky 2007), there are those who believe that immortality is within reach (Futurists); those who believe life expectancy will rise to 100 years or more in this century (Optimists); and those who believe that life expectancy is unlikely to exceed an average of around 85 years in the absence of radical advances in the control of the aging process—and could even decline in developed countries in this century (Realists).” (Carnes and Olshansky 2007:367)
In other work on anti-ageing science I have written on the problems associated with the immortalist / futurist position (Vincent 2006, 2007). In this paper I want to present a radical deconstruction of a highly influential ‘optimist’ position. And point out that parradoxically that this optimist position does much to fuel a great deal of pessimism about population ageing.
There was a great deal of publicity given to a paper by Oeppen and Vaupel (2002) when it was published in the journal ‘Science’ and launched with a press release by the Cambridge University press office. The paper was reported in the major broadsheets and on radio and television with usual sensationalist spin and dire warning that official government projections on population ageing were wrong. In the nature of things population projections as guess about the future are almost certainly are proved wrong, but the dominant interpretation in the media was to suggest that health, pensions and welfare expenditures would become unaffordable. For example: “People are set to live increasingly long lives, and reaching 100 will soon be "commonplace", say experts …"As the cost of pensions spirals there's mounting pressure... to raise the age for retirement” …"The acute problem for society will be how to look after all the older people“ (Heap, BBC Thursday, 9 May, 2002 ); “Health crisis looms as life expectancy soars Average ageing forecasts far too low, say scientists” (Meek, Guardian May 10, 2002). This paper has gone on to be the largely unacknowledged source of alarmist statistics on population ageing in the media and political debate. In this paper I will critique alarmist demography of old age and in particular the selective use of demographic analysis to bolster neo-liberal welfare agendas. To be fair to the authors the original paper is written as an attack on persisteantly over-conservative assumptions about future life expectancy made by demographers and leading international institutions. However, what has happened is that their polemical project became hardened into taken for granted alarmist ‘facts’ about future population ageing.
Oeppen and Vaupel (2002) state that there is a very high linear correlation between female life expectancy in the record-holding country; this has risen at a steady pace of 3 months per year over the last 160 years. Japanese females life expectancy currently 85 it will only take another 45 years to take it (or some other record breaking country) to 100. They argue that all previous attempts to specify maximum life expectancy have failed and conclude there are no limits to life expectancy.
Jim Oeppen and James W.Vaupel (2002) “Broken Limits to Life Expectancy” www.sciencemag.org Scıence 296 10 May 2002 p.1029
The flaws in the argument are manifold. I will look at their method and their data selection and also examin their unstated theoretical assumptions. Their most basic methodology flaw is to rely on mathematic projection without a clear theoretical base and this in effect equates correlation with cause. They assume the high correlation is in itself important and can be used as an explanation in the absence of an explicit theory which links cause and effect. The paper contains no developed theory as to why life expectancy will continue to grow without limitation. Insofar as this point is theorised at all, they refer to continuously occurring gains in medical knowledge. They are, of course, forced to say that they do not believe people will become immortal but without any theoretical underpinning they are unable to comment on when and how their prediction will cease to be valid. The logic of the extrapolation used by Oeppen and Vaupel could also be used to apply to Olympic running speeds. This conclusion would be equally wrong.[i] We could use the same correlation and show, on the basis of Oeppen and Vaupel’s analysis, that life expectancy was zero in 1660.
“Critics of a purely mathematical approach to the study of human life span suggest that the validity of limit hypotheses cannot be ascertained without considering biological evidence on senescence (Carnes, Olshansky, and Grahn 2002). Duration of life is an outcome variable for researchers in the demographic and actuarial sciences. From a biological perspective, however, duration of life is the product of a multidimensional process that involves behvioural, environmental, genetic, and random forces (Carnes and Olshansky 2001; Finch and Kirkwood 2000). As such, there is no defensible basis for the claims, based on purely mathematical models of mortality, that there are no biological or demographic reasons why death cannot decline to zero (Wilmoth 2001) and, in a more cautious formulation of the idea, that life expectancy at birth can rise to 100 years of more (Oeppen and Vaupel 2002).” (Olshansky et al 2002: 504)
Thus there is a necessity to critique the paper from the perspective of other theories and other disciplines. In particular, the Biology of ageing has been undergoing rapid change in the last decade with new findings and new theories. The significance of these findings for the potentiality of devising methods to control, slow or reverse the fundemental ageing process is controversial. However, their is a clear consensus within the bio-gerontological community that the are no current techniques capable of extending the maximal human life span.
Oeppen and Vaupel’s argue that the reasons for continued failure of predictions about future life expectancy is that the trend in life expectancy improvement trailing off as it approached some proposed biological maximum has not materialised. They use that to suggest that the idea of a maximum human life span is a flawed hypothesis and that no maximum life span can be reliably imputed. However, there are others who take a contrary position.
“Today, aging and death are viewed as the inadvertent but inevitable byproducts of the degradation of biological structures and processes that evolved for growth, development, and reproduction rather than for extended operation. These structural and functional constraints exist at every level of biological organization (cells, tissues, organs, and organ systems) within the individual, and their existence imposes practical limits on the life span of individuals and the life expectancy of populations.” Carnes and Olshanski 2002 p.510
In fact biological theories of ageing and the evolutionary basis for age span across species is highly contentious. But the core of Carnes and Olshanski’s argument stands, namely that mathematical extrapolation without reference to an understanding of the biological mechanisms for ageing and the specific components of the gains in life expectancy is an inherently weak argument. There are strong medical and biological reasons for not reallying merely on mathematical extrapolation.
The authors may not be responsable for all the inaccurate interpretations of their argument. However, the approach used in the paper can be criticised for selective use of data. The use of the most extreme longevity at any particular date and projecting that forward almost inevitably slides from being thought of as a projection of an extreme to a prediction about future life expectancy in general. Which is indeed how many popular commentators and policy makers have used these “findings”.
The popular understanding “life expectancy” as ‘how long I will live’ is misleading. It is a statistical measure not how long an individual will live in the future, it is a summary of current mortality. More precisely life expectancies are calculated on current age specific survival rates. It should come with a government warning, life expectancy can go down as well as up; in certain sub-Saharan African countries and for some groups in the former USSR life expectancy has declined. Those older women in Japan who manage to survive being born and weaned in Japan in the 1920’s and 30’s and then the rigours of the 2nd world war will be tough as old boots, the weak ones weeded out and the survivors living on a low calorie diet with known benefits in the older years. On the other hand even the most frail and sickly baby born in the developed world now has much stronger survival possibilities in their first year even perhaps without great promise of longevity. Reports suggest that people with Down’s syndrome and other genetic or congenital disorders have expanded their life expectancy very significantly compared to the rest of the population. But such survivors still live substantial shorter lives on average than the rest of us in Britain. These features suggest that current age specific survival rates are not necessarily a reliable predictor of future longevity. It is at least a possibility that further advances in longevity at the oldest ages, even though advancing at the moment, will be particularly hard to achieve in the future. This is shown by the trends to declining gains in longevity in Japan.
Source: Statistics and
Information Department, Minister's Secretariat, Ministry of Health and Welfare,
Vital Statistics of
take over the ageing race from
Although in other contexts through their work on historical demography Oeppen and Vaupel have shown awareness of the problems of historical demographic data, none of these difficulties surface in the paper. The must be considerable doubt that the bureaucratic and institutional procedures in different countries of a 160 year period consistently and reliably record the same things. Age at death, numbers of live births are two features which are open to cultural and historical variation in interpretation and accuracy in official records. Indeed these figures are most problematic at the youngest and oldest ages. Life tables used for calculating life expectancy make assumptions about the span of the oldest age category. In the past numbers have been small so demographers have been content to collapse the highest ages into a single age band and calculate a probability of death on the basis that all will die within the period. Clearly the resulting probability depends on the width of the age band selected. Normally this is a minor problem not leading to great disparity in life expectancy calculations because numbers are so small. But in the circumstances of the debate over the future of longevity at very old ages these assumptions and their resulting inconsistencies become much more critical.
Data issues at the other end of the life table can also be seen to create problems. Age specific mortality rates in the first year of life are critical to calculating the standard measure of life expectancy at birth. However, in historical and cultural terms what is recorded as a birth (as opposed to other categories such as still birth, miscarriage, or not recorded at all) is very varied. Survival of shorter and shorter term babies in hospitals in contemporary Europe contrasts with regular and thus unreported miscarriages at home in other times and places. Hence compared to contemporary practice there is likely to have been an under-recording of infant mortality rates in the past. Further in many cultures, including Japan, there will be a greater tendency to disfavour female births and less likelihood to record them and strive strenuously to keep them alive. These features suggest that projecting forward on the basis of historically and culturally variable data should be done with caution.
There is some research which suggests that perhaps we should view findings on current increases in life expectancy as a cohort phenomenon not necessarily a feature attributable to future ageing populations. So for example Finch and Crimmins (2004), working from the immunological theory of ageing argue, on the basis of historical data from Sweden, that attention must be paid to cohort effects in population ageing. The ability of people in the twentieth century to provide children with a relatively infection free environment as well as issues about childhood nutrition are an important part of changes in later life mortality experienced as the century progressed. However, influence of this factor would not be replicated by further gains in life expectancy for subsequent cohorts.[ii] A finding which is consistent with the predominant ‘s’ shaped pattern of gains in life expectancy. Other researchers have raised the long term significance in the rise in childhood obesity, particularly linking it through epigenetic processes to increased late life diabetes and increased risk of mortality. These cohort perspectives caution against accepting Oeppen and Vaupel’s implicit modernization perspective with its belief in the continuity of progress.
Within the professional actuarial community and the academic demography community there are technical debates on appropriate methodologies including procedures for predicting future mortality (Willets et al 2004). Demographers and actuaries debate not only about what the most likely future trends will be, but also about the best methodologies for projection. For the actuaries their projections have very specific financial consequences, this tends to mean that although they might try to compete with each other over identifying trends and market opportunities in particular segments of the population, they tend to come together in formulating views about the future national trends. One way in which this is done in the U.K. is through their professional association and the Continuous Mortality Investigation Bureau. In 2004 it consulted on and revised standard tables. One of the key debates in the consultation on those revisions were about the relative merits of using extrapolation techniques as opposed to ‘decomposition’ techniques. The latter involve examining changes in to the probability of specific causes of death and aggregating them into an age specific final figure. Both techniques have their limitations and all predictions are wrong in the long term.
Thus it is clear that the concepts and methods of forecasting future numbers of older people are constructed through social processes including the activities of professional bodies and academic disciplines (Day 1992, de Bruijn, Bart J. 1998). Such demographic concepts “dependency ratio” and “life expectancy” contain embedded within them the issues, perceptions, cultural understanding of the social groups who develop and use them. Some ‘experts’ are keener than others on particular techniques and the need to change longevity predictions (Parsons 1977). It has been suggested that professional actuaries have been more reluctant than academics to accept that there might be no limits to human longevity. It is only possible to fully understand these ‘scientific’ debates about predictions of future longevity if they are located in a context of globalisation and in particular the globalisation of financial markets. The terminology employed in these debates has ideological as well as technical functions (c.f. Ruddick 2003 who makes a parrallel case for views of childhood). We need to understand how these experts and their expertise fit into global society and with who and why their rhetoric finds favour.
A brief examination of the history of the concept “Ageing populations” will illustrate the point. Patrice Bourdelais (1998) in a stimulating paper “The ageing of the population: Relevant question of obsolete notion?” demonstrates the origins and the historical inertia in the way that old age is defined. The threshold of sixty he traces in French demography from the 17th century and demonstrates how the idea of an ageing population calculated by reference to the proportions of population over this age have been a source of concern and debate in many periods. What he is able to demonstrate with historical materials on demographic debates and historical demography is the mismatch between changing social situations and life chances of elderly people and the fixity of categorical ascription into age groups. He shows, for example, how changing fertility rates and life expectancy radically alter the position of the generations within the web of family and kinship, making today’s sixty year olds the core generation of kinship networks.
His approach leads him to suggest that it would be better to think about the changing age structure of the population in ways which take into account the changing meaning of old age. He suggests treating old age as years left rather than a fixed age threshold. This would do away with the population time bombs and enable us to think positively about the success which is represented by people living healthy and fulfilled lives into their seventies and beyond. He calculates for France that the percentage of those 75 and over in 1985 was 6.3%, a similar proportion (6.5%) will be 77 or over in 2005, and in 2040 this proportion (6.8%) will be found for those 82 and over. He suggests that it is as reasonable to consider that the health and social contribution of someone who is 82 in 2040 would not be unlike that of someone who is 77 now rather than consider the increasing proportion of those over 75 as a major social problem.
“To refuse to criticise the notion of an ageing population, to accept this kind of calculation and the fixedness of age categories, is simply an admission that the threshold for old age has not evolved for more than 200 years and will not change for the next fifty, that the starting point of 60 years is immutable. It implies that the significance of age does not evolve historically, that it does not constitute an historical variable. The different ages in life – childhood, adolescence and old age – have inspired many works, but the thresholds defining them appear to have escaped historical development.” (Bourdelais 1998:110-111)
“the scientific pertinence of the notion that a population ‘ages’ can be questioned. Undoubtedly, given the very weak evolution shown in the nineteenth century, this kind of measurement was appropriate when Alfred Sauvey introduced it. But the historian has a duty to demonstrate that during the misleading continuity of categories, often unconsciously accepted, realities have changed. Neither the thresholds nor the connotations have budged, and yet everything else has evolved: the sexagenarians’ health, their place in the generational chain, their economic resources and life styles. The example of commentaries meant to clarify population projections has shown that the fixedness of categories on which the notion of old age is based and that, in turn, reinforce it, leads to a very narrow conception of the future. The underlying hypothesis is that the health and economic and social roles of the various age groups remain constant. Can it be admitted that such an idea has harmful effects?” (Bourdelais 1998:128)
Historically different elites have identified particular demographic changes as a threat, and their ideologies can identify not only the cause but also the moral responsibility for these threats (Vincent 1999). These ideologies, having defined the problem, imply courses of action. Powerful elites select tools to exert control to tackle the perceived crises. If a population is thought to be too high they initiate carefully orchestrated family planning programmes; they may even consider sterilisation campaigns, population expulsion, or gas chambers. Or, if the population is too low, immigration, family allowances, nursery schools and medals for the mothers of the nation might ensure the necessary competitive demographic advantage. In the 20C there have been strong links between demography, eugenics and social engineering (Solway 1995).
We can ask critical questions about the circumstances in which population changes become social issues - who is defining those problems and to what purpose?’ There were concerns in the first part of the 20th century about declines in fertility, loss of population in war and the possibility of falling populations (Mass-Observation 1945). In the thirty years following the second world war - the development decades - which saw unprecedented economic growth and the ending of direct Imperial rule, the major concerns were population growth (Demerath, 1984). Population time-bombs, it was argued, were waiting to blow up economic progress and environmental stability. For example, Alan Finlay’s (1995) “Population Crises: the Mathusian Specter” is exclusively addressed to the issues of the growing size of world population and in particular pollution issues which may result, but simply does not mention population ageing. In the last twenty years the time bomb has become that of an ageing population. Fears about population have shifted from over-population to under-population - too many babies to too many old folk.
“Described as an inexorable ‘growing danger’, the ageing population’s further expansion, we are being warned, is inevitable in the first half of the twenty-first century and will result in numerous economic and social difficulties. Hence, France’s misfortunes might be attributed to the egoistic couple who limit their descendants and, consequently, cause the population to age at the expense of their country’s economic strength and international position.” Bourdelais 1998 p.110
There has been almost a cyclical process whereby different and contrasting demographic concerns come to dominate international political and economic agendas. I would suggest that this changing demographic agenda is best understood as reflecting ideological concerns of dominant elites: economic, military, political and financial (Wallerstein 1979 Walters and Blake 1992). National and global economic elites want labour for their enterprises and customers for their products. In general terms, if demographic expansion promotes economic expansion all well and good, but if rapid growth leads to instability, then concerns emerge. Military elites are concerned about military manpower for themselves and their enemies, (c.f. CIA 2001). Political elites need to sustain state control and thus observe, enumerate and manipulate populations.
The elites who control the multinational firms in the global finance industry have a specific set of interests in the success and expansion in the management of the resources generated by “funded” pension schemes (Clark 2000, Minns 2001, Blackburn 2003), . Jeffery Greenberg, the Chair of the MMC board of directors articulates this interest clearly in a speech to fellow corporate executives in which he summaries the firms prospects:
retirement crisis is something you’re all aware of. In the industrialized
world, there are fewer workers supporting a greater number of retirees. It’s
been a growing trend for the last decade, and it’s accelerating. Over the long
term, the “pay as you go” government pension plans will face ever greater
strains. As a result, the need for private retirement plans in economies around
the world will continue to gain traction. Mercer [an MMC subsidiary] is one of
the leading retirement consulting firms. It operates in 40 countries. We
continue to expand its global business with a combination of internal growth
and acquisitions in a number of countries.” (‘Remarks by Jeffrey W. Greenberg’ Sanford C. Bernstein & Co.
Strategic Decisions Conference
The concerns of these elites are expressed in multinational forums, particularly those charged with the economic co-ordination of global markets. The Organisation for Economic Co-operation and Development for example reflects the dominant international agenda on ageing populations. The OECD (1998) Policy Brief “Maintaining prosperity in an ageing society” states that population ageing could threaten future economic growth and prosperity and suggests that yet greater “reform” of countries’ welfare and financial structures are needed. By way of identifying the critical issues they ask the following questions:
“Will it continue to be possible to share societies’ resources between the working generation and its dependent non-working members in ways that do not give rise to unacceptable societal and inter-generational conflicts?
How can the contribution of older people to society and economic prosperity be enhanced?
How should pension, health and long-term care best be reformed?
Which changes in the financial infrastructure are needed to support the development of funded pension systems?
To what extent will ageing OECD countries be able to improve their well-being through growing trade in goods and services and assets, in particular with younger, faster-growing non-OECD countries?” (OECD 1998:1)
But these questions, from the point of view of the dominant agenda, are rhetorical. The ideology of pension fund capitalism dictates that the answers are obviously that older people should expect less, everyone will have to contribute more towards pensions, people will have to work longer for their pensions, the private sector should be left to manage the funds, and we better rely on the US to make sure the younger states do not step out of line.
It is therefore not surprising that Oeppen and Vaupel’s paper met with such a ready audience from powerful social groups in the global economy and by reiteration became reified into the facts of population ageing. What the paper was able to do was legitimate the interests of those who wished to create market opportunities by dismantling state based PAYG pension scheme by attaching the apparent credibility of numerical demographic science to an explosive statistic. What we should not allow is the political and economic debates about priorities for pension systems and economic support for elderly people to be sidelined by apparently technical arguments about population dynamics. Any demographic theory like any scientific theory is only a good as the assumptions and methods on which it is based.
example another leading
TABLE 2 Change required in demographic, labor force, retirement, and benefit variables to reduce pension expenditures by 10 percent in 2050 (G7 average)
Policy variable Change required
Fertility (2000–2050) Increase of 0.27 births per woman
Net migration rate (2000–2050) Increase of 1.8 net migrants per 1,000 population
Employment ratio (2050) Increase of 9 percenta
Mean age at retirement (2050) Increase of 1.6 years
Pension benefits (2050) Reduction of 10 percent
aThis rise in the employment ratio also causes a 1 percent decline in the pensioner ratio.( Bongaarts 2004:19)
John Bongaarts (2004) “Population Aging and the Rising Cost of Public Pensions” Populatıon And Development Revıew 30(1): 1–23 (March 2004)
But Bongaarts does not consider at least two important factors. Firstly
the social distribution of such costs as increase in retirement age or benefits
such as reduced taxation. Secondly, the paper holds the consequences of
economic growth as constant. There are contrasting theories about the economic
effect of ageing populations. However, throughout the twentieth century pension
growth in the
Blackburn, Robin (2003) Banking on Death or, Investing in Life: The History and Future of Pensions. London:Verso
Bongaarts, John (2004) “Population Aging and the Rising Cost of Public Pensions” Populatıon And Development Revıew 30(1): 1–23 (March 2004)
Bonoli, Giuliano (2003) “Two Worlds of Pension Reform in Western Europe.” Comparative Politics 35(4):399-416.
Bourdelais, Patrice (1998) “The ageing of the population: Relevant question of obsolete notion?” pp. 110-131 in eds. Paul Johnson and Pat Thane Old Age from Antiquity to Post-Modernity. London: Routledge.
Carnes, Bruce A. and S. Jay Olshansky (2007)A Realist View of Aging, Mortality, and Future Longevity Populatıon and Development Revıew 33(2):367–381(June 2007)
Carter, M. N. (2000) “Capital Markets and Occupational Pensions: Opportunities for the United States and Japan” Marshall N. Carter Chairman and CEO, State Street Corporation. Website http://www.us-japan.org/boston/carter.htm.
CIA (2001) Long Term Global Demographic Trends: Reshaping the geo-political landscape. July 2001 obtainable from: http://www.odci.gov/cia/reports/index.html.
Clark, Gordon L. (2000) Pension Fund Capitalism. Oxford: Oxford University Press
Day, Lincoln H. (1992) The Future of Low-Birthrate Populations. London: Routledge.
de Bruijn, Bart J. (1998) Foundations of demographic theory: Choice, process, context. ??: NethurD Publications.
Demerath, Nicolas J. (1984) “World Politics and Population” pp. 33-52 in Susan Strange (ed.) Paths to International Political Economy. London: George Allen & Unwin.
Ermisch, John (1983) The Political Economy of Demographic Change: Causes and implications of population trends in Great Britain. London: Heinemann.
Finlay, Alan (1995) “Population Crises: the Malthusian Specter” pp.152-174 in R.J.Johnson, Peter J.Taylor and Michael J. Watts Geographies of Global Change. Oxford: BlackwellsFinch, Caleb E. and Eileen M. Crimmins (2004) “Inflammatory Exposure and Historical Changes in Human Life-Spans” Scıence 17 September 2004 Vol 305: 1736-1739
Mass-Observation (1945) Britain and Her Birth-Rate. London: John Murray.
Minns, Richard (2001) The Cold War in Welfare: Stock Markets versus Pensions. London: Verso.
Minns, Richard (2003) “Pensions of mass destruction”. Soundings 24:55-61
OECD (1998) Policy Brief “Maintaining prosperity in an ageing society” June 1998. www1.oecd.org/publications/Pol_brief/1999/9907eng.pdf
Oeppen, Jim and James W.Vaupel (2002) “Broken Limits to Life Expectancy” www.sciencemag.org Scıence 296 (No. 5570) 10 May 2002 pp.1029-32??
Olshansky, S. Jay, Carnes, Bruce A. And Jacob Brody (2002) “A biodemographic interpretation of life span.” Population and Devopment Review 28(3):501-13. September 2002.
Parsons, Jack (1977) Population Fallacies Elek Books: London
Ruddick, Sue (2003) “The Politics of Aging: Globalization and the Restructuring of Youth and Childhood” Antipode 35:334-363
Solway, Richard A. (1995) Demography and Degeneration: Eugenics and the Declining Birthrate in Twentieth-Century Britain. London: University of Chapel Hill Press.
Vincent, John A. (1996) ‘Whose afraid of an ageing population?’ Critical Social Policy. 47:3-26.
Vincent, John A. (2006) “Ageing Contested: Anti-ageing Science and the Cultural Construction of Old Age.” Sociology 40,4:681 – 698.
Vincent, John A. (2007). Science and imagery in the ‘war on old age’ Ageing and Society 27(6):1-21.
Wallerstein, Immanuel, (1979) The capitalist world-economy: essays by Immanuel Wallerstein Cambridge: Cambridge University Press.
Walters, Robert S. and David H. Blake (1992) The Politics of Global Economic Relations. 4th Ed. Englewood Cliffs: Prentice-Hall.
Willets, R.C., A.P. Gallop, P.A. Leandro, J.L.C. Lu, A.S. Macdonald, K.A. Miller, S.J. Richards, N. Robjohns, J.P. Ryan and H.R. Water (2004) Longevity in the 21st Century Report Presented to the Faculty of Actuaries, 15 March 2004.
[i] See http://commons.bcit.ca/physics/rjw/pers/womenrun.htm for a hilarious set of projections based on linear extrapolation of 20C Olympic running records.
Finch and Cummins (2004) suggest
“Most explanations of the increase in life expectancy at older ages over history emphasize the importance of medical and public health factors of a particular historical period. We propose that the reduction in lifetime exposure to infectious diseases and other sources of inflammation— a cohort mechanism—has also made an important contribution to the historical decline in old-age mortality. Analysis of birth cohorts across the life-span since 1751 in Sweden reveals strong associations between early-age mortality and subsequent mortality in the same cohorts. We propose that a “cohort morbidity phenotype” represents inflammatory processes that persist from early age into adult life” (abstract)
“The stability of cohort mortality slopes (Gompertz slope), despite the remarkable variability of overall mortality (Fig. 1A) (18), implies that future increases in life expectancy from reduced inflammatory causes may be relatively small, particularly in populations that have had low levels of childhood infection for many decades and now may be approaching a lower limit. These findings from many fields suggest that inflammatory processes that influence the outcomes of aging can be kindled or quenched by exposure to extrinsic infections, inflammatory stimuli, and nutrition. A new theory of human health in life history could emerge from a fuller accounting of inflammatory exposures from gestation to old age.” (conclusion)