Edward's Book Draft

About This Weblog

2008-02-16T06:20:00.000-08:00

This weblog is a personal project to bring together (online, in real time) my thoughts on global-imbalances, demography, fertility and economic growth. Eventually the posts will all form one continuous block which can be read sequentially. In the meantime it is a 'work in progress' with bits of me serving as 'gum and chickenwire' to hold together a bunch of arguments from and links to the pertinent papers.

I hope the outcome will be readable and useful to someone.

Introduction

2008-02-15T19:06:00.000-08:00

We are getting older.

This commonplace is both self-evident - individually we are always that little bit older, each and every day - and surprising - Niger is getting older, Mali is getting older, Somalia is getting older. This is surprising since these are, effectively, among the youngest societies on earth (Niger 15.8 Mali median age 16.35, Somalia median age 17.59). Everyone is aware that Japan is getting older, everyone is aware that Germany is getting older (these are currently the two oldest societies on the planet), but Niger, Mali and Somalia!

In fact, apart from 18 'demographic outliers' as identified in the 2005 United Nations Human Development Report, each and every country on the planet is getting older. Nor is this societal 'ageing' a recent phenomenon, it starts from virually the outset of what has become known as the demographic transition - something which began in many European societies in the late 18th and early 19th centuries. The transition begins with a sudden and sustained drop in mortality, the society becomes 'suddenly young' and after this it is continuous ageing all the way.

So if this work is about ageing, its starting point is that this ageing is not a new or recent phenomenon (the 'discovery of ageing' is of course more recent, but that is another story). The end point? Well there is no end point, as life expectancy continues to push ever onwards and upwards we will all be living longer, and to date there does not seem to be any special biological limit to this process. That is the good news.

The bad news comes with the how. Following the sharp mortality decline which characterises the onset of the demographic transition ageing commences, but historically it has done so at a relatively slow rate. The difficult part of the ageing process as it affects us today is that now proceeding more rapidly, with the evidence suggesting that those societies which began the transition later are ageing even more rapidly. Countries like China and Brazil have experienced sharp declines in their birth rates accompanied by rapid increases in life expectancy. This means that the median age at historically unprecedented rates, and that these societies run the risk of becoming 'old' before they become 'rich'.

The ageing process is also associated in the contemporary world with sizeable changes in the age structure of the population, changes, which given our existing customary boundaries between the ages of work and retirement, are likely to produce a significant increase in elderly dependency ratios across all the OECD countries.

There are however solutions available to address the problems which this rapid ageing will produce. In part it will be the objective of this book to describe and explain the changes which are taking place around us, and in part the aim will be to explain and justify the changes we need to make if contemporary 'ageing' is to become not a threat but a boon and a challenge.

The provisional structure of the book is as follows:

Chap I Nasty Brutish and Short?
Chap II The Second Stage?
Chap III The Discovery of Age Structure
Chap IV The Mysteries of Growth
Chap V The Life Cycle
Chap VI Asymetric Shocks
Chap VI Global Imbalances
Chap VII Remedying the Global Imbalances

Nasty Brutish and Short I

2008-02-15T00:23:00.000-08:00

Nasty Brutish and Short

The 21st century will doubtless bear witness to a great many new and strange phenomena, but somewhere high up on the list of things which are going to define the coming century will surely be the fact that most countries experienced a substantial and sustained ageing in their populations as the century progressed. This phenomenon of population ageing, which has now become something of a commonplace for us, is in fact the result of a twofold process, a general decline in birth rates, and a generalised and substantial increase in the levels of life expectancy.

In the developed world birth rates have long been falling, and are now either hovering around ( one or two countries like the United States and Ireland) or significantly below (the rest of the OECD member states) replacement level. Since there is no consensually agreed theory among demographers which explains why fertility rates have fallen so low the future trajectory of fertility is hard to foresee, but in the immediate future it is clear that for many countries below replacement levels of fertility are likely to remain the norm, and the big question - the billion dollar one in fertility theory terms - is really 'just how low can you go'?

Side by side with this most developing countries are, and will continue for some time to be, in the process of seeing their fertility levels fall steadily (and even dramatically) from TFRs which are initially in the high to moderate band, first to replacement, then to below replacement, and finally to lowest-low fertility and beyond. The case of Iran is instructive here. Having dropped from around 5 to just under 3 between 1989 and 1996, Iran's total fertility rate has continued to decline rapidly, and in the early years of this century dropped below replacement level. data from results from the 2000 Demographic and Health Survey (DHS) for Iran revealed a decline of 32 percent in the TFR between 1996 and 2000, from nearly 3 to around 2 children per woman. The data also shows that reduced fertility is not simply an urban phenomenon but is occurring throughout the country, with some of the most dramatic declines being in the rural areas, where the 1996-2000 decline was some 31% when compared with only an 18% decline in the urban areas (Abbasi-Shavazi, 2002). Data from the Italian statistical agency ISTAT reveal a similar picture with TFRs in the North rising steadily from 1.05 in 1995 to 1.34 in 2005 while in some rural regions fertility fell to very low levels ( Sardinia 1.07, and Molise and Basilicata 1.14, see ISTAT 2006).. Data for East Germany reveal a similar pattern, in that the TFR rapidly converged towards and then went below the already low level of the former Federal Republic. Thus if convergence is taking place (see Wilson, 2001) it is both between and within countries and towards intitially lowest-low fevels of TFR.

At the time of writing some 71 countries (or 43% of the global population) have fertility rates which are in the 2.1 to 5.0 TFR band. These countries are classified by the United Nations as 'intermediate fertility' ones (UN, 2002). Essentially all those countries where TFRs have already fallen below 5-0 may conveniently be described as having initiated their demographic transition, and as such these countries are experiencing steady and sustained fertility declines. It is reasonable to expect that between now and 2050 they will all enter the below replacement fertility group, with the vast majority entering sooner rather than later. As and when they do so they will steadily swell the ranks of the 44 countries who already 'enjoy' this status.

The consequences of such generalised low fertility are not hard to imagine: populations will gradually start to decline and in an increasing number of countries. At the same time our populations will become older, and even a quick glance at any of our garden variety news sources will nowadays normally suffice to find some reference or other to long-term population projections which show truly daunting ageing figures. If, for example, fertility rates in the developed world remain, as they are today, well below replacement, and if we continue to experience improvements in life expectancy at the current rate, then half the population of today's industrialised countries is projected to be over 60 by the time we enter the last quarter of the 21st century. Of course just how 'old' 60 will actually be by the time we get there is an issue in and of itself, and one which forms a significant part of the 'meaty content' of the present ageing debate.

At the same time many developing countries, such as, for example, China, may well find themselves ageing in an especially dramatic fashion, since fertility rates in the third world have been falling very rapidly (UN, 2002) and longevity rates are also rising equally dramatically. So, if the current distribution between working and retirement ages is maintained, in less than two decades China will find itself facing an old-age dependency burden similar to the one which can be found in the older Western European countries today.

The 'ageing phenomenon' itself, of course, willl come as no surprise at all to anyone who has a smattering of knowledge about standard demographic theory, since progressive ageing could arguably be considered to be one of the 'stylised facts' associated with the process which has come to be known as the demographic transition. Now the term demographic transition has normally been used to describe the transition from a demographic regime based on high birth rates, high infant and child mortality rates and low life expectancy to a regime more typified by falling birth, infant and child mortality rates, together with steadily increasing life expectancy. This process has often been thought to occur as a core part of the economic development and modernisation process which carries a country from the condition of being a pre-industrial society to one of being an industrial and then a post-industrial one (Easterlin, 1995, 1983 Caldwell, 1982).

What Is The Demographic Transition?

From the end of World War II and until at least the 1980s, there was probably no research issue that was either mentioned more frequently or debated more passionately in demographic literature than the theories and explanations adavnced in relation to the phenomenon which has become known as the demographic transition.

Transition theory has, right from the start, been characterised by a penchant for a phasal typology, and the original formulation of the theory was no exception to this rule, being presented as it was in terms of a three stage demographic evolution from a regime of high birth and death rates - "a high balance" - to one of low birth and death rates - "a low balance". An intermediate stage of high rates of natural population increase was thought to result from a tendency towards faster declines in death rates than in birth rates. Early pioneers of the theory like Warren Thompson and Frank Notestein, in addition to presenting the basic classification system, also suggested a list of major correlates and causes of fertility decline. These included decreased infant and child mortality, the spread of urbanization, increased costs of raising children, rising parental aspirations, increases in literacy, rises in women's status, a rise in individualism, a decline in religiosity and changes in other cultural factors. Many of these factors have often been included under that wide umbrella which has become known as the modernization processes.

According to the original version of the standard theory, after years of - more or less - homeostatic population drift (often termed the 'Malthusian era'), the start of the transition itself is normally thought to be marked by a sharp and sustained decline in mortality, and in particular by a decline in infant and child mortality. This mortality decline in and of itself produces a large and significant drop in the median age of the society concerned. After this it is, in one sense, up-hill all the way, since normally societies tend to embark on a continuing ageing process, an ageing process which has to date no known end-point. In this sense both the fertility decline associated with the initial or 'first' transition and the more recent one which sees the arrival of below replacement fertility - a process which some have ventured to call the second demographic transition - are intimately related, since the underlying 'driving factors' are undoubedly the same and the whole process is, in its core, an ageing one.

This alone should alert us to at least one difficulty: is it really adviseable to use the expression 'ageing society' in relation to our contemporary developed societies, since if collective ageing is associated with declining fertility and increasing life expectancy, then our societies have long been ageing ones, as indeed they will continue so to be for as far ahead as we can see.

Are There Really Two (Or Should That Be Three) Transitions?

As has been suggested there are those who in addition to breaking the demographic transition down into stages would also break it down into 'transitions'. Insofar as some theorists have gone on to use the expression 'second transition' (Lesthaeghe, 1995, Van de Kaa, 1987) they have normally done so in order to break the entire fertility transition down into two component parts, one which sees fertility decline from its earlier 'higher' pre-modern steady state to the modern, industrial-age near-replacement level, and a second, posterior, transition during which fertility drops from replacement level to one which is substantially below replacement (and in many cases falling as far as what some - Kohler et al, 2002 - have termed 'lowest-low fertility' which they define as lying in the sub-1.3 Total Fertility Rate range). As has previously been argued, there is no sound theoretical or empirical basis to justify making such a radical distinction between a first and a second transition, nonetheless since this terminology is widely used, it will continue to be applied here as a convenient shorthand for describing that package of social and other life course changes which are conventionally associated with the arrival of below replacement fertility.

Stylised Facts Concerning the Modern Demographic Transition

1/. As has been said the key differences between the two 'transitions' (the early pre-industrial transition and the later post-industrial one) is that while in the first case the fertility decline was accompanied by a sharp and continuing drop in infant and child mortality, in the second one it is old-age mortality which declines alongside the continuing fertility decline and it is the life-expectancy of the older-old which rises. This process is sometimes described as the rectangularisation of mortality in reference to the way in which the age-related mortality curve changes during the different stages of the transition, moving as it does from a U shape towards a rectangle as mortality gets gradually compressed at both extremes. The combined impact of the two phases of the transition then is that - after an initial "mortality shock" - all societies enter a process of seemingly continuous ageing. One noteable contemporary consequence of this is the fact that today there are only 18 countries - themselves considered to 'demographic outliers' by the United Nations (and identified as such in the 2005 edition of their Human Development Report) - which are not actually ageing.

So this is the first stylised fact: the demographic transition is a transition from a society with a relatively stationary age structure, and relatively static life expectancy, to one with a gradual and continuous increase in life expectancy and a steadily rising median age. In short, the transitional/post-transition society is an ageing one.

2/. In global terms it is undoubtedly true to say that the twentieth century was the century of life expectancy, since during its course average life expectancy at birth more than doubled - rising from an average of around 30 in 1900 to some 65 years by the end of the century (Riley, 2001). It may well be possible to say the same of the century to come too, since if present trends continue the planetary average life expectancy is projected to further rise to around 81 by the end of the 21st century (Lee, 2003). There is still no consensually agreed explanation for why life expectancy has been increasing in this way. Indeed there is some dispute as to whether in some cases it will now actually begin to decline (the case of obesity in the United States would be but one well-known example of a counter case, declining male life expectancy in Russia would be another, and the Aids epidemic in Africa another (Olshansky et al, 2005, Brainerd and Cutler, 2004, Pelletier, 2004). At the same time there is still no agreement among scholars about whether - as for example James Fries believed (Fries, 1980) - human lifespans have a more or less fixed upper limit or whether a phenomenon known as 'negative senesence' may not be at work (Vaupel et al, 2004). The key question is, as the cohorts which have had heavy disease loads and nutritional deficiencies in their early years pass through their life courses, and life expectancy pushes up towards the previous 'higher limits' whether subsequent cohorts will still exhibit the same upward trend in expectancies. Such uncertaintly not withstanding, and amidst so many potential areas of dispute, one factor does seem to be evident:: the process of increasing life expectancy has been closely associated with increased levels of education and with growing health awareness. In particular almost all studies of the topic highlight the fact that life expectancy and level of education have a strong positive correlation. So significant has the relation between increased life expectancy, increased education and economic growth been across in almost all countries that have experienced a fertility decline, an industrial revolution, and a modern process of economic growth that Columbia University economist Xavier Sali i Martin concluded a recent review of the last 15 years of new growth research with the conclusion that "life expectancy is one of the variables most robustly correlated with growth" (Sali i Martin, 2002). Since as Sam Preston has demonstrated life expectancy is itself correlated with education levels, it is possible that we have here in outline form one important part of the transition transmission mechansism (Preston, 1996)..

Equally the twenty-first century is almost certainly going to be the century when low fertility (meaning by this below replacement fertility) becomes the global norm since as it runs its course the vast majority of our planets population will enter below replacement level fertility regimes. In this sense something important is happening as we are almost certainly fast approaching a turning point in terms of population dynamics, and it is highly probable that global population will peak later in the century, possibly during the last quarter century (UN, 2002).

This then is the the second 'stylised fact' of the transition: the post-transition society has two core characteristics, declining fertility and increasing life-expectency. In fact it is perfectly possible that these two components are themselves interconnected, and this is a possibility which will be extensively explored in the subsequent sections of this book.

3/. The third 'stylised fact' is that while all societies age during the transition, not all age at an equal rate. Differences between societies are evident in terms of the relative rates of fertility decline and the relative rates of increase in life expectancy. Both of these differ substantially from one society to another, and clarifying the factors which lie behind such differences in transition rates and timing is undoubtedly one of the outstanding challenges still facing demographic research.

4/. Another trend has been that those societies who entered the transition earlier generally passed through it at a slower pace, while those who have started their transition later - and in particular in the most recent cases of the drop below replacement fertility - it seems that the process has accelerated considerably (UN, 2002). Indeed it is normally true to say that the later the entrant, the more rapid the subsequent fertility decline and the shorter the generational timescale required to reach the higher life expectancies associated with economic development would seem to be an increasingly valid generalisation. There is possibly an acceleration principle in operation here.

5/. A closer examination of declining fertility quickly brings us face to face with what might be called the fifth stylised fact: about the transition and this is that the fertility transition is normally accompanied by (and in fact often the result of) a steady and continuous increase in woman's mean age at first birth (WMAFB).This state of affairs was often less than evident, since at certain key points in the transition WMAFBs have fallen rather than rising (in England in the late eigtheenth century, during the industrial age following what became known as the second industrial revolution, and more recently after WWII in the context of the so-called 'baby boom' phenomenon). Indeed in the 'second', below replacement, transition this lowering of the MAFB process plays a predominant role, with changes in final parities having a relatively secondary overall importance (Sobotka, 2004). Developed fertility regimes now seem to be characterised by steadily rising MAFBs, and there seems little realistic likelihood that this situation will reverse.

6/ Finally, the increase in WMAFB is normally closely correlated with a steady rise in female education levels, female participation rates in employment and in the number of years of education and training required before entering first employment.. (Elaborate).

Was Life Really So Nasty and Brutish Before the Transition?

Safely armed with the above list of basic 'facts' we are now perhaps in a better position to understand some of the principal characteristics of the demographic transition process itself, but before moving forward to explore this, it may be worth pausing to think a little about life before the transition, life in the pre-transitional society, what was that actually like? How can it best be defined and understood? Was it really, as some have suggested, 'nasty brutish and short'? And if it was so, was this always and everywhere? This question is certainly worth asking ourselves since according to what seems (among economists at least) to be the most widely accepted version our pre-transition reality:

"Until the early 18th century, global population size was relatively static and the lives of the vast majority of people were nasty, brutish, and short.” (Bloom et al 2002).

Or, in the words of Berkeley demographer Ronald Lee (Lee, 2003):

"Before the start of the demographic transition, life was short, births were many, growth was slow and the population was young."

At this point, of course, all the roads lead us back to the work of the very Reverend Thomas Malthus for whom, it will be recalled, slow population growth was no accident. For Malthus population in the long run was held in relatively stable equilibrium with a slowly growing economy by a combination of what he termed 'positive' and 'preventive' checks.

The basis of Malthus's argument is straightforward enough, and really only depends on two fairly simple postulates. The first of these is the evident reality that humans need to eat and thus need food, the production of which in historical societies increases only slowly with time. And secondly there is the idea that 'nature itself' has ordained a passion between the sexes which remains constant through time. This latter 'instinctive urge' should, Malthus thought, lead to a steady stream of children coming continuously online. Since our resource supply can only be improved gradually not all of these children can be be maintained, and herein lies the source of what he termed the positive check.

Now Malthus himself has often been read as suggesting that populations are maintained virtually on the edge of starvation. In fact this is not necessarily the case, either according to the Malthus theory, or in reality. In hunter gatherer societies, for example, it has been estimated that violent deaths, often in warfare, accounted for around 30% of male adult mortality (Coleman 1986). Thus Malthus' mechanism might rather work due to the fact that the increase in population creates additional mortality issues, and one of these is the fact that the increase leads people to take more risks, such as moving farther afield to hunt etc, and in this way the level of 'population attrition' might well turn out to be greater.

Also there is the fact that Malthus himself believed that the greatest mortality problem associated with a limited food supply occured as a result of the fact that those who are under-nourished are more likely to die of infection, especially during epidemics. With more young to fend for, and less nutrition per capita, infants and young children are surely more vulnerable to even the simplest of health insults. This view is really not that far from the opinions of most modern epidemiological theorists and a brief look at the data for death from diahorrea in some third world countries today lamentably still offers a clear and simple illustration of the continuing operation of the process Malthus described.

We don't then need to read Malthus over-simplistically and indeed in some senses his argument here is surprisingly modern: as the Caldwells note, his argument on child nutrition and susceptibility to infection is not at all incompatible with the 20th ventury work of writers like Thomas McKeown and Robert Fogel (Caldwell, and Caldwell, 2003). As I have said Malthus's positive checks were really what we would nowadays call resource constraints: increasing population would inevitably find itself pushing against an agricultural output which was simply unable to keep pace (in a non Boserupian technological environment), and the resulting "misery" and undernourishment leads to a mortality rise which serves, in the long run, to counterbalance the increase itself.

Malthus's preventive checks, on the other hand, were not natural but social in character. They were constituted by socially evolved mechanisms designed to constrain fertility such as, for example, delayed marriage, protracted lactation or pre-modern forms of contraception (although it is important to note that Malthus himself was completely unaware of the possible importance of factors like lactation, or the fact that our reproductive biology may contain its own feedback components which are internalised in what has now become known as 'natural fertility' control: see Henry,1961 Ellison, 2001).

In the case of pre-industrial Europe, as well as in pre-modern foraging societies, Malthus was, in a certain sense, right: population was normally held in a kind of weak equilibrium with its resource environment via a combination of positive and preventive checks. In saying this we need to be cautious in just how we interpret the expression 'weak equilibrium. Robert Fogel, in his Nobel acceptance speech (Fogel, 1993) makes just this point:

"(My) analysis.....points to the misleading nature of the concept of subsistence as Malthus originally used it and as it is still widely used today. Subsistence is not located at the edge of a nutritional cliff, beyond which lies demographic disaster. The evidence outlined in (this) paper implies that rather than one level of subsistence, there are numerous levels at which a population and a food supply can be in equilibrium, in the sense that they can be indefinitely sustained. However, some levels will have smaller people and higher “normal” (non-crisis) mortality than others".

In other words according to the levels of height and the lengths of life-expectancy which are selected-for there are not one, but several, available population subsistence equilibria, and the situation is far from being a deterministic one. Factors well beyond the simple availabilty of the food supply have been at work, and many of these factors were 'natural' in origin (climate, disease) while many of the adaptive responses which evolved in conjunction with these natural ones were social in character (delaying childbirth, increasing terms of lactation).

Nevertheless, equilibrium, in some very general sense, was normally attained. To put this in another way, the resource/population process was a homeostatic one in the very long run. Population tended to osscilate around some general equlibrium level (allowing for height and life expectancy differentials) until weather, disease or political disturbance knocked population strongly away from this equilibrium. At this point the combination of positive and preventive checks would serve as some form of pre-modern 'automatic stabiliser', and the population 'carrying load' would eventually once more be brought back towards its long run equilibrium trend (allowing here for some very gradual forms of social learning and technological change). As as result, for the best part of 10,000 years global population size was remarkably stable, and growth in absolute numbers was slow. (Lee, 1987, 1997; Lee and Anderson, 2002),

However it is important to note that while there is plenty of evidence to suggest the existence of such a clear long-term trend, there is also plenty to support the idea that sizeable fluctuations around the trend occured, and that life expectancy in Europe, for example, varied significantly during the centuries preceding the industrial revolution.

Malthus himself, (like Cantillon before him) was an astute observer of the social customs of his times, and he was far from unaware of the wide variety of social institutions and practices - such as age of marriage, the percentage of women marrying, extra-marital fertility, or contraceptive practices - which had evolved and which served to mediate the operation of what some would call our 'naked biological' reproductive potential in determining the number of children per woman actually born.

In fact Malthus himself was well aware that in Western Europe prior to 1800 the key 'preventive check' in operation was what we now call the European marriage pattern. This expression is normally used to describe marriage customs which seem to have been widely practised during a long period of historic time across a substantial part of Europe, and in particular across that part of Europe which lies to the west of an imaginary frontier running from St. Petersburg to Trieste. The pattern has been found to have been in operation to varying degrees from the late middle ages onwards, and it was still to be found operating in some parts of rural Europe as late as the early twentieth century. The high point of its operation, however, is undoubtedly to be found during the seventeenth century.(Clark, 2005)

Malthus himself devoted a large part of Book II of his Essay to describing the marriage system in Western Europe and how it was characterised by late marriage. According to Malthus, the majority of European men and women delayed marriage until they had the economic resources necessary to support their families at their desired standard of living. In this sense many of the modern explanations for the birth postponment process which is thought to characterise the 'second' transition are not necessarily 'un-Malthusian', although it is important to note and understand the importance of the fact that while, during the Malthusian regime, first birth ages fluctuate and adjust to maintain a given standard of living, in the post-Malthusian regime the rise in first birth ages is secular, and associated with a steadily improving standard of living.

Malthus in fact believed that such delayed marriages prevented individuals from experiencing a substantial reduction in their standard of living on marriage and at the same time prevented the population from growing rapidly enough for positive checks and generalised misery to come into play . In particular Malthus noted that delayed marriage was especially widespread in Norway, Switzerland, and England (1986/1803, Vol. 2:238). He also believed that marriage rates in England were so low that a substantial fraction of English biological reproductive capacity was not being used (1986/1803, Vol. 2:239). Malthus in fact says:

“In a review of the checks to population in the different states of modern Europe, it appears that the positive checks to population have prevailed less, and the preventive checks more, than in ancient times, and in the more uncultivated parts of the world. The destruction occasioned by war has unquestionably abated…And although in the earlier periods of the history of modern Europe, plagues, famines, and mortal epidemics were not infrequent, yet, as civilization and improvement have advanced, both their frequency and their mortality have been greatly reduced, and in some countries they are now almost unknown. This diminution of the positive checks to population, as it has been certainly much greater in proportion than the actual increase of food and population, must necessarily have been accompanied by an increasing operation of the preventive checks; and probably it may be said with truth that, in almost all the more improved countries of modern

Europe, the principal check which at present keeps the population down to the level of the actual means of subsistence is the prudential restraint on marriage” (Malthus, 1986/1830:254).

While Malthus did not actually produce a systematic theory to explain delayed marriage, all the elements for such a theory were already present in his writings. As a true child of his times Malthus naturally believed that the aspirations for a respectable standard of living which he felt lead to the delay were based on foresight, and the ability to defer gratification, and these he argued were associated with the higher levels of civilization then to be found in Europe as compared with the majority of the rest of the world.

Thus he attempted to ground the operation of the mechanisms to delay or forego marriage on the psychologically based aspirations of individual women and men. The search for a reasonable (socially determined) standard of living was the central underpinning to the decision to delay marriage, and this constituted the first element in Malthus's system. Postponing marriage, according to Malthus, requires an unwillingness to be patient with pain and misfortune. Whereas expectations and acceptance of future suffering might lead to early marriage, the hope of enjoying life would cause people to postpone marriage (Malthus, 1986/1803, Vol. 2:59). The hope of bettering one’s self and the fear of misery and being without the necessities of life would be motivations to emphasize the preventive check (Malthus, l986/l803, Vol. 3:453-454).

If the first Malthusian element was hope, the second was foresight, the ability to forsee the difficulties that would inevitably attend an early marriage coupled with the rearing of numerous children (Malthus, 1986/1798: 99). Foresight, according to Malthus, requires the ability to look around one and recognize the distress and poverty that normally becomes the lot of those with large families and insufficient means. It requires a certain awareness of the difficulties associated with trying to rear and support children without first being firmly established economically (Malthus, 1986/1803,Vol. 2:14).

And beyond hope and foresight there lies gratification and our ability to defer it. Delayed marriage, according to Malthus, requires an ability to postpone the immediate comforts and satisfactions which might be thought to be offered by marriage with the expectation of receiving even greater benefits at some unspecified point in the future. By delaying gratification, individuals could assure for themselves and their children the resources for respectability and happiness. The comforts and conveniences of life were even possible with delayed marriage (Malthus, 1986/1830: 251-252).

Not everyone, of course, agreed with Malthus. Some took the Bosnerupian - and essentially more optimistic - view that economic resources had a tendency to grow faster than population did. In a Lecture presented at the University of Oxford in 1828, Nassau Senior argued that if what he termed savage nations were in “a state of habitual poverty and occasional

famine” this was due to their inability to develop their productive resources rather than an excess in the growth of their population. If they had but scanty populations, he argued, this was because their means of subsistence were even

scantier. In contrast in every 'civilized' country he found that “there is now less poverty than is universal in a savage state.” and he took the presence of this higher standard of living as testimony to the fact that “the means of subsistence have a greater tendency to increase than the population” (Senior, 1831: 47-48).

Archibald Alison also came to a similar conclusion, arguing that “the rapidity of increase in poulation is in the inverse ratio of the means which are afforded of maintaining a family in comfort and independence: it is greatest when these means are the least, and least when they are the greatest” (Alison, 1840, Vol. 1: 112).

One noteworthy point in both the Malthusian and non-Malthusian accounts is that it was consumption aspirations which played a key role in the explanation. There was however one subtle but important difference between the two views. On the Malthusian account it is the desire to maintain consumption which drives the preventive checks, whereas Senior, Alison, and others believed that consumption aspirations increased rapidly with economic growth, and that it is the desire to keep raising consumption which produces the fertility restraint. Indeed so closely did they believe that delayed marriage was tied to consumption aspirations that they postulated that economic expansion would ultimately lead to the increased postponement of marriage and a reduction in the rate of population increase. In this sense they are undoubtedly 'modern', and in some sense their work foreshadows the recent 'second transition' literature.

In truth the core idea that consumption aspirations increased with economic resources preceded both Malthus and his contemporary critics. Adam Ferguson writing in 1767 had already advanced the modern idea that what is necessary in life is vague and relative. According to Ferguson what is necessary “is one thing in the opinion of the savage; another in that of the polished citizen: it has a reference to the fancy, and to the habits of living” (Ferguson, 1980/1767: 142).

Ferguson went on to argue that there is apparently no limit to the expansion of consumption aspirations. “No ultimate remedy is applied to this evil, by merely accumulating wealth; for rare and costly materials, whatever these are, continue to be sought; and if silks and pearl are made common, men will begin to covet some new decorations, which the wealthy alone

can procure. If they are indulged in their humour, their demands are repeated: For it is the continual increase of riches, not any measure attained, that keeps the craving imagination at ease” (Ferguson, 1980/1767: 143).

Later Archibald Alison was even to offer an intergenerational mechanism for increasing artificial wants. According to Alison “each succeeding generation is bred up in the habits of indulgence to which the preceding one only attained by the result of many years of successful exertion. The parent who has raised himself from the middling to the higher ranks of life, or from the lower to the middling by a laborious industry, communicates to his children the habits and the wants to which he latterly succeeded. The gratifications which were considered as the highest objects of ambition, or the last step of luxury during the best years of his life, are regarded as mere necessaries by his posterity” (Alison, 1840, Vol. 1: 104).

This is, of course, is precisely that ideational mechanism which now plays such a central role in modern models of fertility decline which regularly refer to the way in which aspirations incorporated in such notions as 'ideal family size' tend to evolve with time and across generations (Lutz et al, 2006, Watkins 1990).

Ferguson argued in the eighteenth century that “while arts improve, and riches increase; while the possession of individuals, or their prospects of gain, come up to their opinion of what is required to settle a family, they enter on its cares with alacrity. But when the possession, however redundant, falls short of the standard, and a fortune supposed sufficient for marriage is attained with difficulty, population is checked, or begins to decline.” (Ferguson, 1980/1767: 142-143).

This is, in essence, the cohort theory of fertility change which has sometimes been offered to explain the transition to below replacement fertility. (Easterlin, 1987, Macunovich, 2000). Alison went even further, writing that the increase in artificial wants was the “great and important change” that provided the “principal counterpoise which Nature has provided to the principle of population. The indulgence of artificial wants is incompatible with a rapid increase of the human species. If the labourer finds himself burdened early in life with a wife and children, he must forego many enjoyments which otherwise would be within his reach…Strong as the principle of population is, experience proves that these prudential considerations, when suffered to develop themselves, are still stronger, and are perfectly sufficient to restrain the rate of human increase” (Alison, 1840, Vol. 1: 109-110).

Awareness of the existence and extension of this deferment process was in fact so general that Richard Jones felt the confidence to write that “this self-restraint is so far exercised that there is no record of the customary age of marriage having at any time, in any country, coincided with the age of puberty. Its strength increases, and its sphere of operation enlarges, with the advance of civilization” (Jones, 1859: 245).

The European Marriage Pattern and The Historical Record

Now according to Clark the European marriage pattern exhibited four main features (Clark, 2005):

1. A late average age of first marriage. Typically between 24 and 26 for women.

2. No control of fertility within marriage.

3. Large numbers of women never married. Typically 10-25 percent, but in some populations and periods the percent unmarried was even higher.

4. Low illegitimacy rates. Typically less than 6 percent of all births were illegitimate, even though the majority of women of reproductive age were unmarried. Illegitimacy rates were as low as 1.5 percent in England in some decades of the seventeenth century. French illegitimacy rates were probably even lower.

These four features are important, and their association not merely incidental, since age at first marriage, use of contraception, marriage and illegitimacy rates may be considered as key indicators, one whose movement tends to define the dynamic evolution of any given fertility regime whatsoever.

Clearly embarking on a marriage and establishing a home constitutes a key life course event in any society, and one which is often fraught with difficulty. In this sense early modern Europe is no exception requiring as it does the resources to establish and maintain a separate household (and this would of course be a by-product of the various European family and kinship systems which were in operation, Voland, 2000), and the resulting age at first marriage for women seems to have been typically comparatively late, normally averaging around 25. In fact a substantial proportion of women never actually married (Flinn, 1981, Livi-Bacci, 2000), and, as a consequence, although fertility was high within marriage, the total fertility rate (TFR) was only a relatively modest four to five births per woman (Livi-Bacci, 2000). In England, for example, the mean age at first marriage for women in the mid seventeenth century was 25.9 years and 17.5% of women never married.

Obviously there were considerable differences between social classes here, while rich (and often aristocratic) families reduced their fertility earlier, and often more markedly, than the rest of the population (Livi Bacci, 1986) this was a reversal of an earlier historic trend. During most of the pre-modern period there was a strong correlation between wealth, probability of marriage, younger age at marriage, and completed fertility (Voland, 2000). However, restricted inheritance and the desire to concentrate wealth limited the reproductive value of noninheriting sons and daughters. Thus, as we enter the modern period and as life expectancy improved and economic structures became saturated, resource holding groups began to delay marriage into the late 30s and early 40s for men and mid-20s for women (Szreter and Garrett, 2000; Voland, 2000).Generally, wealth brought a higher probability of marriage at a younger age, to a younger spouse, and more children. However,

as environments became more saturated, and local resource competition among siblings differentially affected resource-holding families, as opposed to day laborers, in a way which increased the likelihood of dispersal of later-born children (Clarke and Low, 1992; Towner, 1999, 2001; Voland and Dunbar, 1997) since the benefits to resource holders of having an above average number of children was increasingly offset by a more and more intense sibling competition for access to inheritance (Voland, 2000). One extreme example of this kind of parental manipulation of offspring marital opportunities was polyandry, in which a male sibship jointly marries a single woman in order to avoid division of property and labor among competing households of wives and children of brothers (Crook and Crook, 1988; Haddix, 2001).

The other interesting detail about the European marriage pattern is the degree of variability which it exhibited. In England, for example, during the seventeenth century limitations on fertility turned out to be so severe that population even began to fall. In the mid-eighteenth century, average age at first marriage fell, and continued to fall, dropping more or less consistently from a seventeenth century high of 25.9 to a low-point of 23.4 by the end of the first half of the 19th century. Over the same time scale the percentage of women never marrying fell to around 7 percent, while the illegitimacy rate (despite the much smaller fraction of the female population at risk of having an illegitimate child) rose from 1.5% to 6%. (Clark, 2005)

A comparable state of affairs to that which existed in England was also to be found in the Verviers region of what is now Belgium, where the average age at first marriage in 1650-59 was 25.3 for women. This rose to 27.5 by 1700-9, before falling again to 25.9 during the years 1730-39. (Alter, 1988, Desama, 1985).

Such changes, while apparently small, actually have quite profound effects on achieved fertility outcomes. At the 1660 low-point, for example, each woman in England had an average of only 1.9 surviving offspring (a figure which seems almost contemporary), by 1815 this figure had risen to 3 children per woman. The consequence was, of course, that during the years of the Industrial Revolution population rose rapidly in England, from 6.7 m. in 1770 to 17.7 m. in 1850, with the increase being partially a product of the fact that the drop in marriage age meant more children being born and being born earlier in the life course. Of course, at the same time morbidity was falling and life expectancy rising, and this was the other component driving the increase in population size. Thus, other things being equal, what can be seen here is that in the area of demography from small changes big things do grow.

Outside of Europe data on mortality or fertility are, unfortunately, only occasionally available for most countries before the World War II era (Preston, 1980), but what little information we do have suggests that the European experience is not a-typical.

Lee and Cambell (1996), for example, show how in Liaoning, China, during the late eighteenth and early nineteenth centuries, despite high marriage rates, within-marriage fertility was only at two thirds of the level found in pre-industrial Western Europe (also see Lee and Feng, 1999 and Cambell, Feng and Lee 2002). They also found that it took longer after marriage for the first child to be born than it did in Europe, and that women terminated their child-bearing earlier. It is not known with any degree of confidence why fertility within marriage in Liaoning was so low, but as in pre-Industrial western Europe there is no clear indication of contraceptive use, and the most likely explanation is the existence of social customs that resulted in and sustained lower fertility levels.

On the other hand recent anthropological research tends to suggest that most pre-agricultural societies limit fertility in a variety of ways. The numbers of births recorded in the few remaining contemporary hunter-gatherer societies, for example, are far below the limits of what is biologically possible, and,indeed, often not dissimilar to those found in pre-industrial Europe. (Ellison, 2001, Kramer and Boon, 2002).

As has been suggested already,prior to the agricultural revolution it is possible to argue that, via the processes which have come to be known as 'natural fertility' (Henry, 1961), human fertility was homeostatically regulated. Such 'natural fertility' was effectively regulated by a combination of fluctuating first-birth-ages and changes in the distribution of births, and these were regulated by a combination of taboos and social practices and by changes in nutrition and other environmentally related factors. That such mechanisms exist in human populations is hardly surprising since virtually all complex organisms exhibit some sort of flexibility in both age-at-first-reproduction and fertility rates. Natural selection it seems has resulted in the appearance of physiological and psychological mechanisms by which both organisms and individuals adjust fertility onset and fertility rates in relation to changing environmental conditions. (Kaplan& Gangestad, 2004 ).

In the case of pre-agricultural hunter-gatherer society this regulation seems to have been achieved by means of a variety of reproductive strategies - such as, for example, extended lactation, or nutritionally and activity driven fluctuating age at menarche - many of which have the characteristic of being biological responses to a constantly changing external environment (Ellison, 2001). Put in other words such mechanisms might well be described as 'natures contraceptives'.

The view that foragers limit fertility so as to maintains their populations in dynamic equilibrium with available resources has become common currency among anthropologists since the 1970s (Dumond 1975; Hayden 1972, 1986).

One source of critique for the standard 'natural fertility view' originates in the work of Hill and Hurtado and their pithy observation that ‘No natural fertility population yet observed is characterized by zero growth, as would be required over much of our species’ history’ (Hill and Hurtado, 1996: 471). Following this work there has been a greater appreciation of the fact that individual energetic efficiency in resource acquisition and production, rather than total productivity rates or the carrying capacity of the environment, may well play the critical part in determining reproduction rates (Belovsky 1988; Hawkes and O’Connell 1992; Winterhalder et al. 1988). Also the idea that human population history has been characterized not by a series of stepped dynamic equilibria but rather a saw-tooth pattern of periods of rapid growth interrupted by infrequent but serious crashes has become increasingly recognized as an alternative explanation for near-zero growth through much of human prehistory.(Blurton-Jones et al. 1999; Boone and Kessler 1999; Hill and Hurtado 1996: 471–2; Keckler 1997).

In any environment in which humans find themselves, there is typically a wide array of animal and plant food items that could be successfully captured, collected, processed and eaten. And yet, rarely is it the case that human populations capture, collect and consume everything that available. In some contexts, human foragers tend to ignore small mammals, reptiles and birds, while, in others, such prey are pursued and consumed. Plants foods that are relatively time-consuming to collect and process, such as acorns or other seeds, are ignored by some foragers, and routinely collected, processed and consumed by others. The question thus arises as to what kinds of factors actually affect how humans choose which food items to pursue, process and consume?

In an important and systematic study of pre-modern fertility patterns, Campbell and Wood elaborate a cross-cultural tabulation of total fertility rates (TFRs) for 70 forager, horticultural, and intensive agricultural societies basing themselves on the contemporary ethnographic record. Their findings show that there is very little significant difference in TFRs across subsistence regimes (Campbell and Wood, 1988). Hewlett carried out a similar analysis of 40 mobile and sedentary foragers and pastoralists. He found the existence of slightly higher fertility rates among pastoralists, although the difference was not significant (Hewlett, 1991).

Bentley et al. subsequently published an extensive critique and re-evaluation of the Campbell and Wood study, presenting their owncross-cultural comparison of 57 forager, horticultural, and intensive agricultural groups. Using a subset of the Campbell and Wood sample, excluding non-independent cases (ethnic groups that were closely related) and populations with high levels of sterility, they found that intensive agriculturalists had significantly higher fertility rates (Bentley et al 1993b).

Although pre-transitional fertility was often higher in the agricultural societies of the third-world in the twentieth century than it was in the earlier European case, its levels were normally far below the hypothetical biological upper limit for a population, which is normally thought to be around 15 to 17 births per woman (Bongaarts, 1978).

The consensus view amongst economists (see eg Kremer 1993) that, on an aggregate level, population growth has been globally slow over the past millennium is almost certainly correct. This slow growth has, however, also been characterised by a puzzling phenomenon of large and significant variance from the mean, with large swings about the growth path being evident - examples of this would be the stagnation in the fourteenth and seventeenth centuries and a more rapid growth rate in the fifteenth and eighteenth centuries. While exchanges of disease (and techniques for treating them) through exploration and trade may have played some role in this, and variability in relations between pathogens and their human hosts may form another part of the picture (Fridlizius 1984; Perrenoud 1984), global climatic change was possibly at the end of the day the principal driving force (Galloway, 1988, 1987, 1986).

What is the Demographic Transition?

The expression 'demographic transition' has been used here extensively, and a number of stylised characteristics have even been offered, but perhaps it is worth looking just a little bit more closely at this concept, its origins, its history and its current useage. The concept is of course a familiar and widely used one. Perhaps it has become just a little too familiar, producing the sort of familarity that breeds if not contempt then at least a 'taking for granted' of the kind which should make us wary, especially when, as will be explained in more detail below, some of its underlying assumptions are being subjected to be continuous questioning and revision. The concept itself may well, in fact, turn out to be something of a "false friend".

The body of theory which undelies our modern concept of the demographic transition was first advanced by Warren S. Thompson’s in a now classic work , "Population" (Thompson, 1929). Thompson in the now time-honoured fashion broke the process of demographic change down into evolutionary stages according to the various levels levels of the birth, death and natural growth rates. Following Thompson's lead the 'theory proper' was subsequently elaborated by a group of Princeton-based researchers - Kinslay Davis, Frank Notestein and Irene Taeuber - in the years immediately after WWII (Davis, 1945, Notestein, 1945, Taeuber, 1945). The theory was really a huge generalisation based on earlier studies they had carried out of mortality and fertility declines, as such it was clearly a product of its times, and of the data which was then available. In particular this meant the Swedish data, which was among the first to be systematically compiled, and which was to mark the schematisation of the transition from the very start.

Taking Sweden as its prototype, the theory attempted to explain the mortality and fertility changes which accompanied the revolution in living standards and social conditions which followed from the agricultural and industrial revolutions of the late eighteenth and early nineteenth century. As has been repeatedly mentioned thought about the transition itself seems from the very begining characterised by an obsession with breaking it down into phases. In what could fairly be called the moden 'authorised version' the phasal structure is essentially presented by Ronald Lee in the following way:

"The classic demographic transition starts with mortality decline, followed after a time by reduced fertility, leading to an interval of first increased and then decreased population growth and, finally, population aging" (Lee, 2003).

This is a pretty normal typology, encapsulating as it does:

i) A pre-transition 'Malthusian' regime

ii) A mortality decline phase

iii) A reduced fertility phase (itself subdivided into (a) an accelerating population growth sub-phase and (b) a decelerating population growth sub-phase

iv) An ageing phase

As we have already noted, population ageing must start from the moment the fertility decline begins. There are however other problems and many of these have long been known. Lee himself draws our attention to the existence of cases in which fertility declined before mortality, most notably the United States and France, while, as mentioned above, in the UK case fertility rose at the same time as mortality declined. As a consequence of the appearance of such anomolies the theory has, over the years, been subjected to considerable criticism. (See Chesnais 1992 for a good summary of the waxings-and-wanings of the theory across time).

Some critics have even gone so far as to suggest that the demographic transition cannot really be characterised as a theory at all (in any meaningful sense of the term) , and would better be seen as a useful rule-of-thumb generalisation. The prestigous Italian demographer Massimo Livi-Bacci once famoulsy proposed that "we should just destroy all this nonsense of transition theory" (cited in Coale 1994). The essential grounds for much of the criticism has been the discovery through subsequent research that the transition itself has, as mentioned above, followed quite different patterns in different places. In fairness to the original Princeton researchers it should be pointed out here that the real fertility-decline part of the transition only began in earnest in the third world with the arrival of the 1960s, while good collection and analysis of much of the European data only took place in the 1970s and 80s. The failure of the transition to materialise in the third world did, of course lead to ridicule being heaped - often most unjustifiably - on the heads of many demographers. Robert Fogel has recently humourously speculated that the hand of some malevelolent deity or other may be at work since there is a clear rule that demographers run out of patience after about 20 years, if what they theorize would happen doesn't happen, and the deity intervenes so that the expected event - in this case the fertility decline - arrives since it was just about the same time as leading demographers began saying the theory of the demographic transition was dead, that the fertility rate in third world countries, including amongst them many Islamic countries, began to decline rapidly (Fogel, 2005).

As suggested above, one problem case for the theory was immediately presented by France, since the French data show that while mortality began to fall in the early nineteenth century just as it did in Sweden, birth rates commenced a long and steady decline at more or less the same time (and here the difference with the Swedish case couldn't be clearer). Consequently what the earlier theorists had called the second and third phases of the transition coincided. Indeed, no substantial population growth corresponding to that which occured in Sweden took place in France. In England, on the other hand, where mortality likewise began to fall early in the nineteenth century, birth rates were actually rising at the same time (Schofield 1984). England therefore experienced a period of extremely high population growth throughout the 19th century.

As has been indicated above, in order to believe in the explanatory value of the original version of the demographic transition theory it is also important to believe in the idea that in the agrarian society preceding the demographic transition, mortality was always at a high level even if varing substantially from year to year. It is, however, precisely this assumption that has provided a second important difficulty for the original version of the theory. The essential idea was that during the so-called 'Malthusian Regime' mortality was high due to a combination of low living-standards (as population tested the limits of agricultural technology) and recurrent epidemics.

However many historical observers soon started to note that death rates in pre-industrial Europe fluctuated widely across decades and across the centuries. Now as long as the theory was based primarily on the Swedish population data (originating in the mid-eighteenth century) the pre-industrial high-mortality data seemed relatively secure. This situation, however was to change as scholars started to realise that before the mid-eighteenth century the European mortality level was far from stable. Patterns of fluctuating mortality had been demonstrated by a number of scholars, but normally only for individual parishes or institutions, or for regions like southern Sweden (Bengtsson and Oeppen 1993) or northern Italy (Galloway 1994). What really put the cat among the pidgeons for the classical version of transition theory was the surfacing the data on English population dynamics (basically these started arriving from the early 1980s onwards) producing a discrepancy which it was impossible to ignore. In particular Wrigley and Schofield - in a study of population in England from the 1540s to the 1870s - found the existence of a markedly different state of affairs to that which would have been anticpated by the standard theory.(Wrigley and Schofield 1981).

Wrigley and Schofield’s classical data provide evidence which is not entirely consistent with a straightforward working of the original Malthusian mechanism. For quite long periods, notably the early 17th century and in the 19th century, there is - in contradiction with what the Malthusian model would anticipate - a simultaneous increase in both population and the real wages. As a consequence there does not appear to be any strong Malthusian type link between mortality and the standard of living in England at this time, and death rates seem to vary in a way which is hard to correlate precisely with the movement of real wages.

The Swedish demographer Bo Malmberg, after an in-depth examination of Wrigley and Schofield’s data, discovered that during the whole period between 1541 and 1871 the English population underwent several cycles of age structure change. Correlating these changes in age structure with the real wage index it became evident to Malmberg that real wages were high when there were large increases in the 30-64 age group. (Malmberg and Sommestad, 2000) He also found a negative correlation between the share of young adults and the real wage. In addition he found these correlations to be valid across the entire 1541 to 1871 period.

Much as Malmberg's work may throw light on what was actually driving the movements in real wages, the mortality problem would appear to be a much more serious one, since if mortality fluctuations cannot be neatly tied down to changes in living standards then this presents real problem for traditional transition theory, since part of the causal mechanism it seems to rely on is called into question.

It is hard here to overstate the importance of the English population studies in initiating the subsequent re-consideration of transition theory, since what these studies did was reinforce the idea that the norms and institutions regulating birth rates, in both the long and the short run, were sensitive to movements in economic indicators, with the important proviso that they did this in a rather surprising way since, it turned out, it was marriage and not mortality which was the principal regulating factor: when times get harder fewer people would get married, producing a simultaneous rise in the marriage age, and a simaltaneous decline in the fertility rate (the 'guilty party' being the now notorious 'tempo effect', which was only really 'discovered and analysed' by Bongaarts and Feeney at the end of the twentieth century, Bongaarts and Feeney 1998).

Thus the work of Wrigley and Scofield was path-breaking in the sense that their results represented a frontal challenge to two of the basic assumptions contained in the original formulation of demographic transition theory. In the first place mortality levels were not high and stable, and in the second it was births, not deaths, which were influenced by the longer-term movements in the economy. Subsequent Swedish studies, which were conducted in the light of the earlier English work, re-examined demographic and economic co-variance in eighteenth century Sweden and came up with results which pointed in a similar direction to the English ones (Fridlizius 1984). Births and marriages were found to be considerably more sensitive to changes in the economy, both long- and short-term, than mortality (Bengtsson 1993, Galloway 1988), and population changes seemed more influenced by economic developments than by social norms.

As I say above, following the initial mortality decline all societies are effectively ageing, the ageing is continuous, and at the present time it is hard to identify a natural barrier to this process. In this sense the transition doesn't really seem to have an 'end state', and thus can hardly be called a transition, since the word transition seems to imply something. If there is in fact a transition it is one from a society homeostatically balanced around high mortality to one which is pivoted around low and steadily declining mortality.

Having said this, and in fairness to Lee, what may be meant by ageing is a society with a comparatively high proportion of dependent elderly. On this view the initial mortality decline creates a dependency ratio which is considerably higher than that in the earlier agricultural society. This 'imbalance' takes many years to correct as fertility rates remain high and societies slowly recover the earlier ratios. But equilibrium is not recovered, and dependency ratios once more start to rise, this time amongst the elderly population. So this is what many may mean by ageing societies: societies where elderly dependency ratios rise (and continue to increase) above a certain notional level.

This way of looking at things has a certain validity, but it does beg one very important - indeed possibly critical from a policy perspective - question: just what do we mean by 'old'. The expression, like the terms modern and post-modern is a deceptive one, since it gives the impression of veing carved eternally in time, when in fact it is, of course, an extraordinarily relative one. To give one illustrative example, one populist Turkish politician got himself elected on a promise to introduce male pensions from the age of 43 and female ones from the age of 39 (something which, of course, resulted in the worst pension's crisis in history). He presumeably thought that 43 was 'old' and those who voted him into power evidently agreed. What we consider to be old is a socially defined (and hence relative) concept. It will hold different values at different times, and as life expectancy reaches ever higher limits we can expect our definition to adjust accordingly. This topic however, will have to await a later stage in the argument to receive the elaboration which it deserves. Simply consider this a foretaste.

Whatever the ultimate verdict on the validity of the phases schema, it should be noted that societies which enter the transition later tend to pass through it at an ever increasing rate. This if we take the mortality decline component we can see that gains in life expectancy have occured in the twentieth century in developing countries at rates which are rapidly by historical standards. In India, life expectancy rose from around 24 years in 1920 to 62 years today (a gain of 0.48 years per calendar year over 80 years), while in China, life expectancy rose from 41 in 1950–1955 to 70 in 1995–1999, (a gain of 0.65 years per year over 45 years.(Lee 2003) Fertility transitions since World War II have typically been more rapid than those for the developed countries, with fertility reaching replacement in 20 to 30 years after onset for those countries that have now completed the transition. Fertility transitions in east Asia have been particularly early and rapid, while those in south Asia and Latin America have been slower in starting but now seem to be accelerating (Casterline, 2001, United Nations Population Division, 2003).

Cohort-Based Mortality

2008-02-14T10:15:00.000-08:00

As has been outlined in the previous section, belief in the explanatory value of the traditional version of the demographic transition theory hinged, in part,on the idea that in the agrarian society which preceded the transition mortality was always and everywhere at a relatively high level, even if did, as a matter of fact, vary not insubstantially from one year to another. It is, however, precisely this assumption that has furnished one of the most important difficulties for the original version of the theory, since historical observers have long been able to confirm the proposition that death rates in pre-industrial Europe fluctuated widely from one decade to another over a time-span of centuries.

The McKeown Hypothesis

Awareness of this reality lead researchers to begin to explore a variety of issues associated with the social and medical determinants of health, and to examine how these might impact on mortality. The standard account of the 'great mortality decline' which accompanied the transition can in many ways be traced back to the work of Thomas McKeown (1976, 1979). McKeown argued that economic growth and better nutrition were the fundamental causes of the remarkable improvements in population health which were seen in Western Europe from the late 18th century onwards. His criticism of earlier views was based on a study of cause-specific mortality in England and Wales from 1838 to 1947, where he found that two-thirds of the mortality decline was due to a reduction in infectious diseases. In later work he also analyzed mortality rates and economic development for a wider group of countries and argued that, contrary to what was at the time the received opinion, medical advances had little direct influence on health before the breakthrough which followed the generalised use of sulphonamides and antibiotics in the 1930s and 40s. According to McKeon, up to this time the only disease that active medical treatment had been able to cure was diphtheria, and this even this was achieved via the use of an antitoxin whose existence itself only dated back to around 1900. Thus only a very small part of the pre-twentieth century mortality decline could , on McKeon's account, be credited to active medical intervention, and even where medical advance was pertinent - as in the case of the diphtheria antitoxin - it could hardly claim to have played a significant part in the mortality conquest, since the disease was in fact already in remission by the time the antitoxin was developed and widely available.

Addressing the earlier part of the mortality decline - from the late 18th to the mid 19th centuries - McKeown made the seemingly valid point that medical influence on one of the key 'killer diseases', smallpox, was not essentially attributable to modern medicine since vaccination (despite being available from the end of the eighteenth century) was not widely used in England until after 1840 when it became freely available at public expense. Gunnar Fridlizius has also drawn similar conclusions regarding the evolution of the decline in Sweden (Fridlizius 1984).

According to McKeown, while an improvement of personal hygiene may have had some effect on mortality in England and Wales after about 1880, when a decline in intestinal infections coincided with substantial improvements in water supply and sewage control, a change which must surely have reduced the incidence of waterborne infections. Since changes in mortality from intestinal infections constituted only a small part of the general mortality decline prior to 1870 the general reduction must have been in large part due to factors other than improved personal hygiene and equally those public health measures which were introduced must have had little impact on the great mortality decline prior to 1870 and only a partial effect thereafter.

In fact long before the time of McKeown researchers had been aware of the existence of apparently 'spontaneous' changes in mortality, and of the changes in population size that these seemed to produce (Helleiner,1957, Chambers, 1972, Fridlizius, 1984, Schofield,1984). In particular Helleiner argued, based on Western European data, that population sizes had fluctuated substantially with a marked increase occuring from the mid-eleventh century until the late thirteenth century and then again from the mid-fifteenth century to the end of the sixteenth century.

The population growth which occured in the eighteenth century was therefore not unique, except insofar as mortality started its decline from a higher level and then continued for longer than it had done previously. McKeown's argument is essentially that since the eighteenth century decline forms an initial (integrated) part of what was later to become the great decline, it is not plausible simply to view it as being a spontaneous decline, one which was due, for example, to a remittance in the virulence of pathogens. McKeown's strong argument seemed to lie in the fact that once started this decline then continued across the entirety of the two following centuries. In explaining the decline McKeown himself focused on the role played by nutrition, and argued that the nutritional improvement determined not only declining impact of infectious diseases from 1848 onwards, but also the initial phase of the decline which began in the late 18th century.

His intransigence over the idea of a common nutritional causal component, and his consequent 'condensation' of what some consider to be two phenomena into one is really the by-product of his laudable desire to find a single common explanation for the entire process of the great mortality decline.

Inflamation, and the Thrifty Phenotype Hypothesis

The Mckeown thesis, however, has not worn well with time (see, for example, Emily Grundy's review: The McKeown debate: time for burial, Grundy, 2005) and ideas which were to seriously challenge what had to all intent and purpose had become the "Mckeown consensus" were not long in appearing. The death knell was perhaps first tolled among what might seem a rather surprising congregation: the practitioners of modern epidemiology. In his seminal work "Mothers, Babies, and Disease in Later Life" (Barker, 1994), the epidemiologist David Barker brought together under one roof what was at the time a growing body of medical evidence which drew attention the apparent importance of fetal nutrition in the subsequent health of the mature adult. The "Barker hypothesis" - which is sometimes referred to as the "womb with a view" hypothesis (Deaton, 2005) - is essentially constituted by the idea that events in the womb have long-lasting effects on health throughout the entire lifespan, and especially effects on health outcomes that only express themselves later in the life course. More technically put, nutritional insults in utero, which prevent the foetus developing to its full potential, or which produce an adaptation ill suited to the external environment which the individual will ultimately encounter, may cause a selective abandonment of function, an abandonment which disfavors or disables precisely those features in the organism whose primary function is to prevent disease in late life beyond the normal reproductive span. While the hypothesis itself is still somewhat controvesial, evidence in support of it has been mounting in recent years.

In this context Gabrielle Dobblhammer and James Vaupel (Dobblhammer and Vaupel 2000, Dobblhammer, 2002), for example, have shown that life expectancy at 50 varies seasonally depending on the month of birth. According to their findings, in the northern hemisphere, 50 year olds in the cohorts studied who were born in the months of October and November (to mothers who perhaps had had better access to cheap and plentiful fresh fruits, vegetables, and eggs through most of their pregnancy) could expect to live about three-quarters of a year longer than those born in the spring.

In the southern hemisphere, the same seasonal pattern was found to occur, although with a six-month shift in the timing of the effect, while those born in the Northern hemisphere who die in the South (European immigrants to Australia, for example) continue to display the Northern pattern.

Other similar evidence comes from the Dutch famine of 1943, follow-up studies on which seem to support the claim that nutritional deficits in pregnancy have long-term consequences for, for example, obesity, with deficits in the first trimester of pregnancy predicting later adiposity, and deficits in the third trimester inhibiting it. There have even been findings which relate subsequent behavioural disorders and schizophrenia to early prenatal nutrition in this context.(Susser and Lin, 1992, Neugebauer, Hoek, Ravelli et al, 1998, 1999, Rosebooma, 2000, Susser, 1999, van der Zee, 1998).

Indeed, explosions of obesity and associated diseases (adult-onset diabetes, heart disease, and so on) have often been found to come come close on the heels of a loosening of nutritional constraints, when those whose parents were undernourished, who were themselves undernourished in utero, move into an environment in which food is plentiful and heavy manual work is no longer the norm. One study found, for example, that in the black township of Khayelitsha near Cape Town in South Africa, more than a half of the adult women had body-mass indexes above 30 (Case and Deaton, 2005).

One clear implication of the Barker hypothesis is that the health of the adult may be a function of birth timing: the when of birth. This has given rise to what has come to be called cohort analysis in epidemiological research, a line of investigation which attempts to explore how disease incidence and life expectancy vary across cohorts in differing (and especially extreme) environmental settings.

The Cohort Thesis and the Great Mortality Decline

As has been said, what later became known as 'the great mortality decline' began in Western and Northern Europe around the middle of the eighteenth century (in some cases possibly a little earlier), levelled off slightly in the mid-nineteenth century, and then continued an inexorable downward course. During this time life expectancy at birth rose from around 35 years to more than 70. In several countries the increase in life expectancy was, indeed, truly spectacular: For the earliest cohorts to have been systematically studied - Sweden 1751, France 1806, England 1841, and Switzerland 1876 - cohort life expectancy at birth was initially very low: 34 years in Sweden, 38 years in France, 42 years in England, and 45 years in Switzerland. By the 1899 cohort, however, life expectancy had jumped to 55 years in Sweden, 56 years in Switzerland, 53 years in England, and 50 years in France.

In these initial (prototypical) European cases mortality began to decline somewhere between 50 and 150 years before the arrival of the industrial revolution in each country and in any event significantly before living standards started their long monotonic upward movement. Also life expectancy normally tended to start to rise some 100 to 150 years before marital fertility started its long-term decline. It should be noted that there are important exceptions to this 'rule'. In England the decline in death rates started around the same time as the initiation of the industrial revolution, during, as it happened, a time of falling real wages, while in France, fertility started declining at about the same time as mortality did, and both of these changes again took place well before living standards started to improve. This having been said, it is noteworthy how the timing of the great mortality decline was strikingly similar across the countries of Western and Northern Europe despite the not insignificant differences in their respective levels of economic and social development. It was also often surprisingly simultaneous in different regions of the same country despite again the large differences in the internal economic development of the countries concerned. (Bengtsson, 2001).

Initially the decline was characterised by a dramatic and sustained decline in infant and child mortality, however later in the nineteenth century improvements in adult mortality also began to occur. Adult and old age mortality had in fact started to decline slowly right from the beginning of the nineteenth century, and possibly even earlier for England. But this decline became much more pronounced in the latter part of the nineteenth century and accelerated after World War I along with mortality at all other ages. The decline then slowed for adults and the elderly around 1950 but from the 1970s onwards it has once more continued apace (Crimmins and Finch, 2006).

In the context of the 'great mortality decline' when we talk about cohort-based health factors what we are normally talking about are factors which affect only certain generational groups, factors which may nonetheless may have longlasting effects on the lifetime health of those groups themselves.

In fact in terms of the great mortality decline 'cohort analysis' is essentially concerned with with two factors, improvements in nutrition and living conditions during the pre-birth foetal period and in early childhood, and the disease environment present during pregnancy and the early life years of a child. Both of these factors may, through their subsequent impact on health, be associated with longer term changes in life expectancy.

As suggested above, in the literature it is possible to identify two types of cohort-related explanations for the great mortality decline:

(1) increased nutritional intake during the foetal stage and/or early years of life, and

(2) decreased 'effort' during the foetal stage or early childhood being required to fight disease either on the part of the mother or of the child, or both.

In each case these factors operate not only through their impact on short term mortality but through their longer run effects on the health of the individual. One possible mechanism for this process may be via the imact of early life events on the rate of growth of the individual and how this affects long run health. (Mangel and Munch, 2005, Gluckman, Hanson and Spencer, 2005, Metcalfe and Monaghan, 2003). Certainly laboratory studies on rodents have found that severe caloric restriction retards growth (resulting in a small bodied adult) but also lengthens lifespan, which might be thought to suggest that fast growth can have negative impacts on subsequent mortality and lifespan (Rollo, 2002; Metcalfe and Monaghan, 2003). Calorie restriction of rats at young ages has also been found to have a tendancy to slow down growth rates and to lead to short adult stature, even when food becomes abundant later in the juvenile period (Shanley and Kirkwood, 2000).

Epidemiologists and demographers of an earlier generation, and who studied the modern mortality decline during the 1920s and 30s, were already aware of this 'early life history' possibility (Derrick 1927, Kermack et al 1934). They noticed that mortality for children declined much earlier than mortality for adults, and that each succeeding generation seemed to carry with it the same relative mortality from childhood though to old age. Distinguishing here between what are called period and what we have termed cohort effects, if period effects (that is environmentally significant imacts on health like more clement weather, or better nutrition, or rising living standards, operating across a given time period) were the dominant factor in the decline, then the heath consequences of these effects should be found to be evenly distributed between both young and old. If, however, this change is found to be asymmetric with one group showing a different pattern from the other, then there are arguably reasonable prima facie grounds for suspecting that cohort factors may be at work, and this in fact was the conclusion these early researchers began to draw.

In more recent times studies have continued to confirm the impact of cohort membership on health and mortality. Sam Preston and Etienne van de Walle , for example, in their study of urban France, and Gunnar Fridlizius, who examined the Swedish case, found such effects to be significant (Preston and van de Walle, 1978, Fridlizius, 1989). But in the realm of modern economic theory, and in it's interaction with economic history, there can be little doubt that if there has been one scholar who has done than any other to advance our understanding of how the cohort hypothesis might play a central role not only in epidemiological research, but also in our understanding of the process of modern economic growth, it has been Robert Fogel (Fogel, 1993, 1996, 2004).

In support of his thesis that cohort factors play a decisive role in the process of long term improvement in life expectancy Fogel used final heights as a proxy measure of net nutrition and health during childhood. Height is seen by Fogel as a cohort related measure of health, while weight and body mass index are seen as life-period measures (Fogel, 1996). On the Fogel account, individuals who, as a consequence of having had well-nourished and healthy mothers, were well nourished during the foetal stage experience lower death-risk during infancy. If they are well nourished and healthy their cells and organs develop better, they attain a greater height and tend to have a longer life.

Since here it is net and not gross nutrition that determines health and height, there is no direct link between gross nutrition during childhood, or GDP, and heights. This is because improvements in health and height may be the result of either better nutrition, or of reduced claims on health due to the impact of disease, or, of course, of both of these. Thus a decline in the prevalence of smallpox, for example, has a positive effect on heights and on the length of the life span, everything else being equal. One problem that immediately presents itself in this line of research is how to evaluate the extent to which the improvement in height and health is due to diet, as opposed to being due to lower claims from disease. Calculating diets for pre-modern populations is a difficult task (Fogel,1996), and it is even more difficult to calculate disease claims. Still, historical records do show similarities between trends in height and GDP (Fogel 1994, 1996), which suggested to Fogel at least that the trend in disease claims may have been been of lesser importance.

Now if Fogel is right here, then one important immediate consequence, and one that is central to his entire argument, is the absence of any single determinate equilibrium between food supply, population heights and population numbers: the relationship is characterised, in fact, by the esistence of multiple equilibria (Fogel 1994). Undernourishment, whether a result of low or badly-composed food intake, or a consequence of an increased disease claim, may rather lead to a stunting of height or weight and a higher incidence of illness, disease and mortality in later life as opposed to any notable increase in direct and immediate mortality. The one-to-one relationship (or period link) between economic output and mortality is thus much weaker than Malthus appears to have believed, at least on Fogel's account.

Now body size has received a good deal of attention in life history analysis (Roff, 1992, Stearns, 1992), and between species, body size is found to correlate with a number of life history traits, including mortality rates. In general large species, including humans, tend to have lower mortality rates and longer lifespans (Harvey and Zammuto, 1985; Gaillard et al., 1989). As a large bodied mammal, we have relatively low mortality and relatively long lives (though our lifespan seems to be proportionately longer than would be predicted by referring to our body size alone: Hill and Kaplan, 1999; Hill et al., 2001). Within species however the relationship between size and mortality is less clear-cut, since large size may offer some advantages, such as protection from predators, but these advantages do not come without cost, since, for example, there are greater nutrient requirements involved in maintaining a large body (Blanckenhorn, 2000). One complicating factor, as we have noted, is the speed of growth experienced during childhood, which is correlated with final adult height but may also have implications for mortality in adulthood.

The relationship between adult size (height) and mortality in humans has been extensively studied. Changes in height have been shown to correlate with mortality trends in both the US and the UK, with life expectancy appearing to rise with average height (Floud et al., 1990; Fogel, 1993), and being taller has been found to correlate with a lower mortality rate (Marmot et al., 1984; Waaler, 1984), but the situation may not be as straightforward as it appears to be. There is evidence that while the incidence of some causes of death, such as cardio-vascular and respiratory disease, are inversely related to height; others, such as reproductive cancers, increase in frequency with height (Barker et al., 1990; Leon et al., 1995; Smith et al., 2000; Song et al., 2003). There is therefore some debate as to whether being taller is as beneficial as it is sometimes thought to be (Samaras et al., 2003).

It is also not clear what the exact relationship is between measures of body condition and mortality. As in the case of height, there has been a good deal of research has into how exactly BMI interacts with mortality. The relationship is normally thought to be non-linear (Wienpahl et al., 1990; Rissanen et al., 1991; Laara and Rantakallio, 1996; Yuan et al., 1998; Engeland et al., 2003; Kuriyama et al., 2004). Individuals with low BMI experience high mortality rates, but those with high BMI do too.

Being short, on the other hand may also be considered to be an indicator of early life conditions, however when it comes to analysis, correlations are one thing, and explanatory mechanisms another. In this context indirect support for a modified variant of the Fogel hypothesis has come more recently from the work of Caleb Finch and Eileen Crimmins (Finch and Cribbins, 2004a, Crimmins and Finch, 2006). Finch and Crimmins advance the general proposition that a 'cohort morbidity phenotype' may serve as a representative of the inflammatory processes (disease claims) that persist from early age into adult life. Specifically Finch and Crimmins propose the hypothesis that decreased inflammation experienced during early life, which is associated with improved infant and child health, led directly to the subsequent decrease in morbidity and mortality resulting from chronic conditions found in old age. They point out that, for example, later life risk of heart attack and stroke is known to be correlated with serum levels of inflammatory proteins such as C-reactive protein (CRP). At the individual level, CRP levels are also correlated with the number of seropositivities to common pathogens, a relationship which tends to indicate a history of prior infections. Furthermore, drugs with anti-inflammatory properties (nonsteroidal anti-inflammatory drugs, statins etc) have been found to reduce the risk of vascular events and even possibly Alzheimer's disease. This kind of evidence may be read as implying the existence of links between levels of inflammation and major chronic conditions which are important in old age, and thus between exposure to infectious disease in early life and health in old age.

Now if we seek to apply these known correlations to the course of the great mortality decline, the early Swedish example assumes,due to its systematic character, considerable importance, and it is hardly surprising that Finch and Crimmins have recourse to the detailed, micro-level, work of Bengtsson and Lindström, as well as other earlier work based on aggregate data conducted in a UK context by William Kermack and the pioneering work (again using aggregated data) of HB Jones for Sweden (Bengtsson and Lindström, 2003, Kermack et al, 1934, Jones, 1956).

So, using the longer data series that is available today for Sweden (infant mortality data was not available to Jones for cohorts which had been born before 1895) Finch and Crimmins have updated Jones’ earlier work, detailing age-specific mortality rates for five birth cohorts in the years between 1751 and 1940. They find that mortality at any given age across the lifespan drops steadily across successive cohorts. Cohorts with lower young-age mortality also have lower mortality at any given age in later life, and this is entirely consistent with an earlier (and very interesting) Jones hypothesis to the effect that “the physiological age of each new generation is remaining more youthful at the same chronological age”.

As Finch and Crimmins emphasise the historical demography of Sweden offers an unparralled possibility of deriving unique mortality profiles across the entire life span, starting with the years immediately prior to the industrial revolution (when mortality was, of course, high) and following each cohort across the entire life course from birth to old age. Taking this data as their starting point they proceed to examine age-specific mortality trajectories for Sweden from 1751 right through to 1940, and find that the data offer support to the hypothesis that old-age mortality declined in a cohort and not a period fashion across all ages. In so doing they develop two points which were essentially already hinted at in the earlier work of Kermack et al. and Jones:

(i) that the historical mortality decline among the old and young begins in the same cohort, and

(ii) that infant mortality has a stronger relationship to later-life mortality than does mortality in subsequent childhood years.

They also conclude that declines in mortality after age 70 tend to lag about 70 years behind those for infants. When they relate childhood mortality to later-age mortality for Swedish birth cohorts born in the 177-year period from 1751 to 1927, they find strong relationships between rates of childhood mortality and mortality for cohort survivors in old age, indeed they found that most of the identified variance in cohort mortality was explicable in terms of mortality before the age of 10. Moreover, they also found that the annualized effect of each childhood year on old-age mortality was three times as great for infant mortality as it was for mortality in subsequent childhood years.

Based on this study of the Swedish data they go on to argue that the inflammatory-infection and Barker fetal-nutrition hypotheses may be seen not as competing but rather as complementary hypotheses, in that they jointly link the two mechanisms of morbidity between early and later life. As they argue, even well-fed babies are vulnerable to rampant infections, and infections alone can cause malnutrition and later dietary deficiencies. Childhood diarrheas, for example, impair cardiac muscle synthesis, and this could underlie the associations which have been found between infant diarrhea and later cardiovascular disease . As they suggest slowed infant growth under the Barker hypothesis could in part be consequent to infections that cause inflammatory responses as well as impairing nutrient absorption.

In a similar vein Bengtsson and Lindstrom, using longitudinal data, and following individual cases rather than relying on grouped aggregate data - a limitation which had characterised the earlier work of Kermack, Fridlizius, and others - have studied historical Swedish cohorts to test both the nutritional and the inflammation hypotheses. They did this by examining the effects of food prices and the disease load at the time of birth on subsequent old age mortality during the years 1766–1894. They conclude that the level of infection among infants was a stronger influence than food availability on later-life mortality and life expectancy. In particular they identify problems leading to the impairment of the respiratory mechanism as the principal source of this influence. (Bengtsson and Lindström, 2000, 2003)

Barbi and Vaupel - in a rejoinder to Finch and Crimmins (Barbi and Vaupel, 2004) - have objected to their findings on the ground that the most recent analyses of mortality patterns over age and time have revealed that period effects are generally more important than cohort ones in explaining mortality decline at the older ages and that, in fact contemporary demographic and epidemiological studies tend to suggest that the cohort effect is at best modest.. In defence of their position they cite, for example, the Danish twin studies which indicate that less than 10% of the variation in how long these twins live is attributable to variation in shared health conditions early in life ( McGue et al, 1993, Herskind et al, 1996). In particular they point out that, in developed countries at least, progress in reducing old-age mortality accelerated around 1950 and accelerated even further around 1970, doing so simultaneously at all older ages.

Finch and Cribbins (2004b) have responded to this by pointing out that since their analysis explicitly excludes modern birth cohorts, members of which have benefited from immunizations and the use of antibiotics, many of the points made by Barbi and Vaupel have limited validity in the context of their argument. They specifically hypothesize that inflammation associated with vascular disease and cancer (the incidence of which is attenuated by modern drugs with anti-inflammatory activities) is the strongest connective link between early and later cohort mortality and that such cohort inflammatory mechanisms are most active when mortality from infections is high. As childhood infection has decreased due to immunization, public health advances, and the use of antibiotics, early inflammatory exposure has had much less impact on cohort old-age mortality for the modern cohorts.

What we seem to have here are two interrelated, but distinct phenomena, the pre-1950 cohort-related effects of decreased childhood inflammation on average life expectancies, and the post 1950 improvement in mortality

rates at the older ages. At this level the arguments of Finch & Crimmins and Barbi & Vaupel are entirely compatible, with the former having a high degree of relevance to the pre-1950 situation, and the latter to the post 1950 one.Now analytically these processes are really quite distinct, as is the economic interpretation which can be given to each of them. Basically, following Finch and Crimmins, we might say that a predominance of cohort influences characterise the first stage, whilst (following Vaupel) period (or environmental and health care) influences characterise the

second one. It also raises the rather interesting point about whether Jones, when he observed that "the physiological age of each new generation is remaining more youthful at the same chronological age" may not have been looking at cohorts which came from the first stage of mortality decline, and not cohorts which form part of the elderly expectancy improvement we are currently seeing. If this is so the implications will be important.

As indicated above the age-specific mortality trajectories from 1751 to 1940 used in the Finch and Cribbins work strongly suggest that old-age mortality declined in a cohort, and not a period, fashion. The mortality trends at age 70 in any given calendar year, or the period mortality trend in old age, do not resemble the trend for the younger age groups. In fact they find that, following an initial rise after 1751, mortality declines first became significant in the Swedish 1791 cohort, and this at both the young and the older ages for that cohort. Period mortality, on the other hand, first declined significantly among the old in the years from 1861 to 1870, years, of course, which correspond to the very cohort in which the onset of the decline was first observed. Again, generally speaking child mortality trends correlate less with old-age mortality trends in the same year (period effect) than with mortality trends seven decades later (Finch and Crimmins, 2004b).

Barbi and Vaupel's critique has not, however, been completely barren, and it has forced Finch and Crimmins to sharpen and clarify their argument considerably. Hence, in a second, and subsequent, work on the same core topic (Crimmins and Finch, 2006) , where they extend their analysis to France, England and Switzerland, they are at pains to point out that they:

"focus exclusively on cohorts born before the 20th century, when levels of infection were high, but before smoking, a major inflammatory stimulus, became popular. Most importantly, these cohorts entered adulthood before general childhood immunizations and before antibiotics. The inflammatory mechanisms that we describe can only work when mortality from infection is high; once childhood infection is low, it can no longer be a factor in explaining old-age

trends."

In fact in their second paper Crimmins and Finch produce some really intriguing cohort-relative life-expectancy data. For the earliest cohorts they study (Sweden 1751, France 1806, England 1841, and Switzerland 1876) cohort life expectancy at birth was low: 34 years in Sweden, 38 years in France, 42 years in England, and 45 years in Switzerland. By the time we get to the 1899 cohort, however, life expectancy has jumped to 55 years in Sweden, 56 years in Switzerland, 53 years in England, and 50 years in France. In both cases the comparatively low life expectancies imply that all the cohorts (both the earlier and the later ones) were exposed to the then highly prevalent infections.

Confirmation of these Crimmins and Finch findings comes in more recent work from Tommy Bengsston (in association this time with Göran Broström). Bengtsson and Broström once more develop a methodology to try to test whether or not events which occur during the subsequent life course may mediate the effects of early-life factors on later life mortality, and in particular whether the degree of access to land in adult life plays any kind of role (Bengtsson and Broström, 2006). Bengtsson and Broström find no support for the null hypothesis that the influence of disease load in the first year of life is not permanent throughout life but is moderated by an individual’s socioeconomic condition later in life (and more specifically at age 50 years). They find that those who (according to the land-wealth criteria they use) could be considered economically unsuccessful by the time they reached 50 did not suffer more from the damage caused by the first year of life disease load than those who had done relatively well (economically speaking) and who had attained or retained access to land. They also find that those who were exposed to a heavy disease load in the first year of life, and who survived to be 50, had an estimated remaining median life expectancy of about two years less than those who were born in years with low to moderately high infant mortality: This is indeed an intersting finding as it makes exposure to infection during the birth year a more important determinant of later life health than sex or socio-economic status.

Similar results showing links between early infections and late-life health have also been found in the case of Union Army veterans in the United States using data from the current Health and Retirement sample (Costa, 2000).

As Crimmins and Finch also point out , maternal infections, including influenza, malaria, and tuberculosis, were common in Europe and the United States well into the 20th century (Riley, 2001). Babies of mothers with infections are known to reveal elevated inflammatory markers and retarded uterine growth (Moorman et al, 1999) and Crimmins and Finch even specualte that suboptimal adult female health may transgenerationally transmit the imprints of infections and inflammation as well as malnutrition while increasing the risk of smaller babies with lowered resistance to environmental pathogens. This additional path is not developed in the Barker hypothesis and is consistent with observations that improved infant mortality lags a generation behind the decline in adult mortality (Kermack et al, 1934).

Now at this point the argument becomes truly interesting. Fogel himself has recently proposed that a ‘‘techno-physiological revolution’’ increased energy available for growth and improved resistance to infection through a dual mechanism which both improved food production and at the same time lead to higher incomes which enabling an ongoing revolution in living conditions (Fogel, 2004). The Fogel hypothesis has been thought to present the difficulty that increases in height did not always follow increases in income and nutrition; and certainly not in the way his theory would anticipate. Height has even been found to have decreased during some periods of improving income in early industrial cities (Flood et al, 1990). However modifying (or blending) the Fogel hypothesis with the work of Crimmins and Finch it can be argued that a decrease in infections and ensuing inflammation had the potential to increase height independently of improved food intake, thus making the joint hypothesis far more compatible with the observed evidence.

Also, as Hillard Kaplan would argue, 'Life is an energy harvesting process'. More specifically this process is characterised by a series of trade-offs, of which the most important are those between growth, maintenance and reproduction. Since energy used for one purpose cannot be used for another (the ‘principal of allocation’), much of what has come to be known as life history theory is concerned with the functioning and impact of such energetic trade-offs. As Kaplan says:

"Organisms capture energy (resources) from the environment. Their capture rate (or income) determines their energy budget. At any point in time, they can "spend" income on three different activities. Through growth, organisms can increase their energy capture rates in the future, thus increasing their future fertility. For this reason, organisms typically have a juvenile phase in which fertility is zero until they reach a size at which some allocation to reproduction

increases fitness more than growth. Through maintainance, organisms repair somatic tissue, allocate energy to immune function, engage in further energy production, and so on. Through reproduction, organisms replicate genes. How organisms solve this energetic tradeoff shapes their life histories."

(Kaplan and Gangestad, 2004)

Mixing this further with an old idea of Lionel Robbins that 'economics is the science which studies human behavior as a relationship between given ends and scarce means which have alternative uses.' we can begin to see just how such trade-offs may have important implications.

Bengtsson and Broström, for example, find that:

"Children born in years with very high disease load, face more than 90 percent higher mortality than the others after controlling for all the covariates included in the model"

Well lets think about this for a moment, and lets think about it in the context of the behavioural relationship between scarce means and conflicting demands, in the context of Kaplans trichotomy between growth, reproduction and maintenance and lets go back in order to do so to Vaupel's original objection to Finch and Crimmins. Which was

that:

"while Finch and Crimmins hypothesize that decreased inflammation during early life has led directly to a decrease in morbidity and mortality resulting from chronic conditions in old age.......demographic and epidemiological studies suggest that the effect is modest".

Well, as we have noted, this leads Finch and Crimmins to respond to Vaupel with a much sharper version of their argument. In particular the qualify their argument by stating:

"Our analysis excludes modern birth cohorts, individuals of which have benefited from immunizations and the use of antibiotics....(while)...The comment by Barbi and Vaupel incorrectly implies that death rates among the elderly in developed countries declined only after 1950.....As childhood infection decreases because of immunization, public

health advances, and antibiotics, the early inflammatory exposure has much less impact on cohort old-age mortality."

So what we have here are two interrelated, but distinct phenomena, the pre-1950 cohort-related effects of decreased childhood inflammation on average life expectancies, and the post 1950 improvement in mortality levels in the older ages.

Crimmins and Finch in fact clearly spell this in their 2006 PNAS piece:

"We focus exclusively on cohorts born before the 20th century, when levels of infection were high, but before smoking, a major inflammatory stimulus, became popular. Most importantly, these cohorts entered adulthood before general childhood immunizations and before antibiotics. The inflammatory mechanisms that we describe can only work when mortality from infection is high; once childhood infection is low, it can no longer be a factor in explaining old-age trends."

So inflammation is largely a pre-1950 issue (in the Swedish, but not of course, in the current developing world, context) and this is where things get, frankly interesting, especially if we think about Bengsston's finding that children born in years with a low disease load experience around 10% of the mortality exposure of children born in the high disease load years.

These (low disease load) children, not only survive in greater numbers, they also live longer, healthier (and hence logically more productive) lives. Now lets think of this in terms of Kaplan's tripartite trade off. And in terms of economics. And in terms of his embodied capital model.

Firstly the low disease-load years mean the mums need to invest less energy in reproduction, since more children survive. That immediately frees off more energy for growth and maintenance. But, since the children are healthier there is less expenditure on maintenance, or, what amounts to the same thing, the investment in maintenance is more

cost effective.

Then there is growth, and let's think here in terms of economic growth, since as Xavi Sala i Martin nicely points out:

"The relation between most measures of human capital and economic growth is weak. Some measures of health, however, (such as life expectancy) are robustly correlated with growth" (Sala i Martin, 2002)

Now, if we go back to Kaplan we should easily be able to see why this relation between growth and "growth" (which was also to some extent evident to Fogel) should be so.

Kaplan estimates that no child in any society is ever really self-sufficient till the age of around 20. Now in the low disease-load years, getting the individual child to 20, not only involves less maintenance energy, it also produces an individual with say 35 productive years out in front of them instead of. say, none, or at least considerably less than 35. The productive impact of this has to be enormous. Of course this productive impact can only be realised within a technological and institutional context that makes such realisation possible, but in the presence of this we seem to have here a huge increasing-returns type mechanism which can help explain why demographic processes are much more important to economic development and growth than has been hitherto modelled.

This also has very important implications for those contemporary societies where diahorrea and malaria etc are still huge killers, and might give us some indication of how societies which are still caught in this health trap will be able to grow once they break out. The finding is also important since it indicates that such a demographic 'dividend' is only possible in the cases of societies where child-health related inflammation is still an issue, and thus tells us

relatively little about the economic outloook for those societies where the major increases in life expectancy come from improving the outlook in the older age groups.

Is There An End State?

The sum total of everything which has gone before is that the fall in mortality which preceded the industrial revolution may be much better seen not as the start of something new, but as the end of something old. There was, of course, something new to follow (in the shape of better public hygiene, and later imporved medical intervention), but that something "new" did not come onstream until well into the nineteenth century, when general improvements in conditions of life, in the form of better diet, better housing, improved hygiene, better child care and better sanitary systems in the towns, effectively prevented a posterior mortality increase, an increase which had unfailingly taken place following all earlier periods of enduring mortality reductions.

This leads us to one rather obvious and uncomfortable conclusion: all those economic growth models which have predicated the rise of the modern 'growth era' on a fall in mortality consequent to the technology revolution which accompanied the industrial one may in fact have the causal arrows pointing the wrong way.

In the account of Galor and Weil (2000), for example, growing population, through its assumed effect on the growth rate of skill-biased technological progress, causes the rate of return to human capital accumulation to increase. This ultimately leads to sustained growth in per capita income. Jones (1999), argues that increasing returns to accumulable factors (usable knowledge and labour) cause growth rates of population and technological progress to accelerate over time, and eventually, it is this which permits an escape from the Malthusian stagnation. The reality, as we have seen, is more complex, and both the 'weakly-Malthusian' initial state, and the low-fertility, increasing-life-expectancy end state seem to be full of surprises both interms of their implications for the initial demographic transition theory, and for the economic growth theory explanations which have rested on it.

As I have already emphasised, following the initial mortality decline which marks the onset of the transition all societies are effectively ageing. This ageing is a continuous process, and at the present time it is hard to identify an indestructible natural barrier which stands in its way. In this sense the transition doesn't really seem to have an 'end state', and thus can hardly be called a transition, since the word transition seems to imply a movement from something to something. If, in fact, there is a transition it is one from a society homeostatically balanced around high mortality to one which is pivoted around declining fertility, declining mortality, and ever-increasing life expectancy. As Lutz emphasises we don't yet know if there is any lower bound to fertility, and as Vaupel suggests there is now no good reason to assume that life expectancy has any natural upper limit..

Having said all this, and in fairness to Ronald Lee and others who use the expression, what may be meant by the process of 'population ageing' may well be a society with a comparatively high proportion of dependent elderly, as indicated by a conventionally determined life-course anchor point, such as the retirement age.Following the Lee account the initial mortality decline creates a child dependency ratio which is considerably higher than that in the earlier agricultural society. This 'imbalance' takes many years to correct as fertility rates remain high and societies slowly recover the earlier ratios. But equilibrium is not restored, and, after an initial 'sweet demographic period' (which may, as we have seen really be a 'sweet immuniological period', dependency ratios once more start to rise, only this time the rise is amongst the elderly population. This transition is rooted in the structure of the human life history and its mortality representation, with the disease load being attacked asymmetrically, initially in the earlier years, then in the later ones.In fact, what many authors may mean when they talk of ageing societies are societies where elderly dependency ratios rise (and continue to increase) above a certain notional level, and this dependency increase is, in the longer run of things, simply the historical footprint and shadow of the earlier, initial, mortality decline (in other words what we may have is one single 'great mortality decline', or transition, otherwise known as the rectangularisation of human mortality).

The traditional demographic transition way of looking at the mortality decline does, of course, have a certain validity, but it does beg one very important - indeed possibly from a policy perspective critical - question: just what do we mean by 'old'. This expression, like similar socially relative terms - 'modern' and`post-modern' would be good examples - is a deceptive one, since it gives the impression of being carved eternally in time, when in fact it is, of course, extraordinarily relative to our life course and our life history evolution.. To give a simple illustrative example, a populist Turkish politician famously got himself elected during the early 1990s on the promise of introducing comprehensive male pensions starting at the age of 43, with female entitlement starting at the even more 'tender' age of 39 (a policy decision which, of course, resulted in one of the worst pension's crises in world history). The politician in question presumeably considered that the age of 43 was 'old', and those who voted him into power evidently agreed with him. The point really is that what we consider to be old is a socially defined (and hence relative) concept. It will hold different values at different times. In the Turkey of the 1990s life expectancy was not especially high when compared with that which which may now be anticipated in contemporary developed societies, an indeed similar, if not so spectacular, examples of the Turkish definition are to be found littered around the history of the third world. They correspond to an earlier, and rapidly transforming, shape associated with the population pyramids.

However as modern life expectancy breaches ever higher limits we can expect our definition of 'old' to increasingly adjust itself upwards accordingly, and in general we should probably keep our fingers crossed that Jones was right when he surmised - back in 1956 - that “the physiological age of each new generation is remaining more youthful at the same chronological age”.

Whatever the ultimate verdict on the validity or utility of the transition phases schema, we should not leave the topic without noting one last thing: those societies which enter their transition process later tend to pass through it at an ever increasing rate. In the case of the mortality decline component of the transition we can see that gains in life expectancy occured in the twentieth century in developing countries at rates which were very rapid by historical standards. In India, life expectancy rose from around 24 in 1920 to the contemporary level of 62 (a gain of 0.48 years per calendar year over 80 the 80 years in question), while in China, life expectancy rose from 41 in 1950–1955 to 70 in 1995–1999, (a gain of 0.65 years per year over 45 years.(Lee 2003) (Find some other data here, this is less than useless as an illusrtation, to myself, Edward). Like wise fertility transitions since World War II have typically been more rapid than those which occured in the nineteenth century, with fertility reaching replacement level in 20 to 30 years post onset, and then continuing to fall steadily, and apparently inexorably, in the direction of lowest-low fertility.. Fertility transitions in East Asia were particularly early and notably rapid, while those in South Asia and Latin America were slower in starting but now seem to be accelerating very rapidly (Casterline, 2001, United Nations Population Division, 2003).

References

Abbasi-Shavazi, Mohammad Jalal, 2002. Recent changes and the future of fertility in Iran, UN, Paper presented at Completing Fertility Transition, Experts Meeting, New York.

Alison, Archibald. 1840. The Principles of Population. Vol. 1. Edinburgh: William Blackwood and Sons.

Alter. G, 1988. Family and the female life course. The women of Verviers, Belgium, 1849-1880, Madison.

Barbi, Elisabetta and James W. Vaupel, 2004, Comment on "Inflammatory Exposure and Historical Changes in Human Life-Spans", Science 17 June 2005: Vol. 308. no. 5729, p. 1743

Barker, D. J. P., ed. 1992. Fetal and Infant Origins of Adult Disease. London: British Medical Journal.

Barker, D. J. P. (1994), Mothers, babies and disease in later life. London : British Medical Journal

Bengtsson, T. 2001. Mortality: The Great Historical Decline. International Encyclopedia of the Social and Behavioral Sciences, Elsevier Science Ltd, pp.10079–84.

Bengtsson, T (1993): A Re-Interpretation of Population Trends and Cycles in England, France and Sweden, 1751-1860, in Histoire & Mesure, VIII-1/2, 1993.

Bengtsson, T. and Göran Broström, 2006, Old-Age Mortality in a Life-Course Perspective. Southern Sweden, 1829–1894. Paper presented at the Annual Meeting of the Population Association of America, Session 49: The Impact of Childhood on Adult Health, Los Angeles, 30 March – 1 April, 2006,

Bengtsson, T. and Martin Lindström, 2003, Airborne infectious diseases during infancy and mortality in later life in southern Sweden, 1766–1894, International Journal of Epidemiology 2003;32:286-294

Bengtsson, T. and Martin Lindström, 2000. Childhood Misery and Disease in Later Life: The Effects on Mortality in Old Age of Hazards Experienced in Early Life, Southern Sweden, 1760–1894. Population Studies, 263–77.

Bengtsson T, Oeppen J 1993 A reconstruction of the population of Scania 1650–1760. Lund Papers in Economic History 32

Bentley, Gillian, T. Goldberg, and G. Jasienska. 1993a. The fertility of agricultural and non-agricultural societies. Population Studies 47:269–81.

Bentley, Gillian, G. Jasienska, and T. Goldberg. 1993b. Is the fertility of agriculturalists higher than that of nonagriculturalists? current anthropology 34:778–85.

Blanckenhorn, W. U., 2000, The evolution of body size: What keeps organisms small? Quarterly Review of Biology 75, 385-407.

Bloom, D. E., D. Canning and J. Sevilla (2002). The Demographic Dividend: A New Perspective on the Economic Consequences of Population Change. Santa Monica, CA, RAND.

Bongaarts, J. and G. Feeney. 1998. “On the quantum and tempo of fertility,” Population and Development Review 24(2): 271-291.

Boone, J. L. and Kessler, K. L. 1999. More status or more children? Social status, fertility reduction, and long-term Fitness. Evolution and Human Behavior, 20: 257–77.

Brainerd, E and David M. Cutler, 2004. Autopsy on an Empire: Understanding Mortality in Russia and the Former Soviet Union, NBER working paper, WP 10868, Cambridge, Ma

Caldwell J C 1982 Theory of Fertility Decline. Academic Press, London

Caldwell J, and B Caldwell, 2003. Pretransitional population control and equilibrium, in Population Studies 57(2):199-215, Cambridge, United Kingdom

Campbell, K. L. and J. W. Wood. 1988. "Fertility in traditional societies: Social and biological determinants" in Natural human fertility. Edited by P. Diggory, S. Teper, and M. Potts, pp. 39–69. London: Macmillan

Cambell, Cameron, Wang Feng and James Lee, 2002, Pre-Transition Fertility in China :Old Wine in New Bottles, Population and Development Review. December, 2002: 735-750.

Case, Anne, and Angus Deaton, 2005. “Health and wealth among the poor: India and South Africa compared,” American Economic Review, 95(2), May. (Papers and proceedings) May 2005.

Casterline, John. 2001. “The Pace of Fertility Transition: National Patterns in the Second Half of the Twentieth Century,” in Global Fertility Transition, a supplement to Volume 27 of Population and Development Review. Rodolfo Bulatao and John Casterline eds. New York: Population Council, pp. 17–52.

Chambers, J. D. (1972), Population, economy and society in pre-industrial England. London: Oxford University Press.

Chesnais, J-C (1992), The Demographic Transition. Stages, patterns, and Economic Implications. Clarendon Press. Oxford.

Clarke, A.L. and B.S. Low 1992 Ecological correlates of human dispersal in 19th century Sweden. Animal Behavior

44:677-693.

Clark, Gregory.The Economics of the Ascent of Man: A Brief Economic History of the World. Princeton University Press (2005)

Coale, Ansley J. 1994. Conference on European Fertility Bellagio - July 1968. Office of Population Research Working Paper 94-2, Princeton University.

Coale, Ansley J. and Roy Treadway. 1986. “A Summary of the Changing Distribution of Overall Fertility, Marital Fertility, and the Proportion Married in the Provinces of Europe,” in The Decline of Fertility in Europe. Ansley J. Coale and Susan Cotts Watkins, eds. Princeton: Princeton University Press, pp. 31– 181.

Coleman, David. Population regulation: a long-range view. In: The state of population theory: forward from Malthus, edited by David Coleman and Roger Schofield. 1986. 14-41 pp. Basil Blackwell: New York,

Costa, D. L., 2002. Changing Chronic Disease Rates and Long-term Declines in Functional Limitation Among Older Men. Demography. February 2002. 39(1): 119-138.

Crimmins, E and Caleb Finch, 2006. Infection, Inflammation, Height, and Longevity, Proceedings of the National Association for the Advancement of Science, January 10, 2006, vol 103, no 2, 498-503

Crook, J.H., and S.J. Crook 1988 Tibetan polyandry: Problems of adaptation and fitness. Pp. 97-114 in Human Reproductive Behavior: A Darwinian Perspective. L. Betzig, M. Borgerhoff Mulder, and P. Turke, eds. Cambridge: Cambridge University Press.

Davis, K. (1945), “The world demographic transition”, Annals of the American academy of political and social sciences, 237, pp. 1-11.

Deaton, Angus (2005), The Great Escape: A Review Essay on Fogel’s The Escape from Hunger and Premature Death, 1700–2100, Journal of economic Literature, Forthcoming

Derrick, V. P. A. (1927), ”Observation on (1) error on age on the population statistics of England and Wales and (2) the changes of mortality indicated by the national records”, Journal of the Institute of Actuaries, 58.

Desama, Claude. 1985. Population et révolution industrielle. Evolution des structures démographiques à Verviers dans la première moitié du 19e siècle, Paris, Les Belles-Lettres.

Doblhammer, Gabriele, “Differences in Lifespan by Month of Birth for the United States: The Impact of Early Life Events and Conditions on Late Life Mortality,” Max Planck Institute for Demographic Research, Working Paper No. 2002–19 (May.), 2002.

Doblhammer, Gabriele and Vaupel, James W., “Lifespan Depends on Month of Birth,”

Proceedings of the National Academy of Sciences, 2001, 98, pp. 2934–9.

Dumond, D. 1975. The limitation of human population: a natural history. Science, 187: 713–21.

Easterlin, Richard, 1995, "Industrial Revolution and Mortality Revolution: Two of a Kind?" Journal of Evolutionary Economics 5:4 (December), pp. 393-408.

Easterlin, R.A. 1987. Birth and Fortune. The Impact of Numbers on Personal Welfare. New York: Basic Books, 2nd edition.

Easterlin, RA (1983), "Modernisation and Fertility: A Critical Essay." In RA Bulatao and RD Lee (eds),. Determinants of Fertility in Developing Countries, Vol. II. New York: Academic Press, pp. 562-586.

Elo, I. T. and Preston, S. H. 1992. Effects of Early-Life Conditions on Adult Mortality: A Review. Population Index, 58(2): 86–212.

Ellison, Peter T. 2003. Energetics and Reproductive Effort, American Journal of Human Biology 15:342–351

Ellison, Peter T. 2001. On Fertile Ground, A Natural History of Human Reproduction, Cambridge, Harvard University Press.

Engeland, A., Bjorge, T., Selmer, R. M., Tverdal, A., 2003, Height and body mass index in relation to total mortality. Epidemiology 14, 293-299

Ferguson, Adam. 1980/1767. An Essay on the History of Civil Society. New Brunswick, U.S.A: Transaction Books.

Finch, Caleb E. and Eileen M. Crimmins, 2004a, Inflammatory Exposure and Historical Changes in Human Life-Spans, Science 17 September 2004: Vol. 305. no. 5691, pp. 1736 - 1739

Finch, Caleb E. and Eileen M. Crimmins, 2004b, Response to Comment on "Inflammatory Exposure and Historical Changes in Human Life-Spans" Science 17 June 2005: Vol. 308. no. 5729, p. 1743

Floud, R., Wachter, K. W., Gregory, A., 1990, Height, Health and History: Nutritional Status in the United Kingdom, 1750-1980 (Cambridge University Press, Cambridge).

Fogel, R. W. (1993), ”New sources and new techniques for the study of secular trends in nutritional status, health, mortality, and the process of aging”, Historical Methods, vol. 26, no. 1, pp. 5-43.

Fogel, R. W. (1996), ”The escape from hunger and premature death 1700-2100: Europe, America and the Third World”, the 1996 Ellen McArthur Lectures, Cambridge.

Fogel, R. W. (2004) The Escape from Hunger and Premature Death, 1700–2100: Europe, America, and the Third World (Cambridge Univ. Press, New York), p. 31.

Fogel, R. W. (2005). Reconsidering Expectations of Economic Growth After World War II from the Perspective of 2004. IMF Staff Papers, Vol 52, Washington

Fridlizus, G. (1984), ”The mortality decline in the first phase of the demographic transition: Swedish experiences”, in Bengtsson, T., Fridlizius, G. and Ohlsson, R., eds., Preindustrial population change: The mortality decline and short-term population movements. Lund: Amqvist & Wiksell International, pp. 71-114.

Fridlizius, G. (1989), ”The deformation of cohorts: Nineteenth century mortality decline in a generational perspective”, Scandinavian economic history review, vol. 37, no. 3

Fries, J. F. 1980. “Aging, natural death, and the compression of morbidity,” New England Journal of Medicine 202(3): 130–135.

Gaillard, J. M., Pontier, D., Allaine, D., Lebreton, J. D., Trouvilliez, J., Clobert, J., 1989, An analysis of demographic tactics in birds and mammals. Oikos 56, 59-76.

Galloway, Patrick R, 1986. Long-Term Fluctuations in Climate and Population in the Preindustrial Era , Population and Development Review, Vol. 12, No. 1 (Mar., 1986) , pp. 1-24

Galloway, Patrick R, 1987. Differentials in demographic responses to annual price variations in pre-revolutionary France A comparison of rich and poor areas in rouen, 1681 to 1787, European Journal of Population, Volume 2, Numbers 3-4 , 269 - 305

Galloway, P (1988), "Basic patterns in annual variaitons in fertility, nuptiality, mortality, and prices in pre-industrial Europé", Population Studies, vol 42, no 2.

Galloway, Patrick R., 1988, "Prussia vital registration and census data description tables 1849 to 1914 using Kreise and cities >20,000 as units of analysis," MS, Department of Demography, University of California, Berkeley.

Galloway, Patrick R., 1991, "Fertility decline in Prussia: a regional analysis," presented by Lee and Hammel at the International Union for the Scientific Study of , 1994 Population Seminar on Old and New Methods in Historical Demography.

Galloway P R 1994 A reconstruction of the population of North Italy from 1650 to 1881 using annual inverse projection with comparison to England, France, and Sweden. European Journal of Population 10: 223–74

Haddix, K. 2001 When polyandry falls apart: Leaving your wife and your brothers. Evolution and Human Behavior 22:47-60

Hamilton, W. D. 1966. “The moulding of senescence by natural selection,” Journal of Theoretical Biology 12(1): 12–45.

Harvey, P. H., Zammuto, R. M., 1985, Patterns of mortality and age at first reproduction in natural populations of mammals. Nature 315, 319-320.

Hayden, B. 1972. Population control among hunter-gatherers. World Archaeology, 4: 205–21

Gluckman, Peter D, Mark A. Hanson and Hamish G. Spencer, 2005. Predictive adaptive responses and human evolution,Trends in Ecology and Evolution Vol.20 No.10 October 2005

Grundy, E, 2002 The McKeown Debate: Time for Burial, International Journal of Epidemiology 2005 34(3):529-533

Haddix, K. 2001 When polyandry falls apart: Leaving your wife and your brothers. Evolution and Human Behavior 22:47-60.

Helleiner, K. F. (1957), ”The vital revolution reconsidered”, Canadian journal of economics and political science, vol. 23, pp. 1-9.

Herskind, A.M. Matthew McGue, Niels V. Holm, Thorkild I. A. Sørensen, Bent Harvald, James W. Vaupel, 1996. The heritability of human longevity: a population-based study of 2872 Danish twin pairs born 1870-1900, Human Genetics,

Volume 97, Number 3 , 319 - 323

L. Henry 1961, ‘Some data on natural fertility’, Eugenics Quarterly, 8: 81-91.

Hewlett, Barry S. 1991. Demography and childcare in preindustrial societies. Journal of Anthropological Research 47: 1–39.

Hill, K. and Hurtado, M. 1996. Ache Life History: The Ecology and Demography of a Foraging People. Chicago: Aldine de Gruyter.

ISTAT, 2006, Indicatori demografici Anno 2005, Istituto Mazionale di stadistica, Rome, Italy

Jasienska G. 2001. Why energy expenditure causes reproductive suppression in women: an evolutionary

and bioenergetic perspective. In: Ellison PT, editor. Reproductive ecology and human evolution. New York: Aldine de Gruyter.

Jones, H.B, 1956, A special consideration of the aging process, disease, and life expectancy, Adv Biol Med Phys. 1956;4:281-337

Kaplan, Hillard S. & Steven W. Gangestad. 2004. Life History and Evolutionary Psychology, University of New Mexico, mimeo

Keckler, C. N. W. 1997. Catastrophic mortality in simulations of forager age-at death: where did all the humans go? In Integrating Archaeological Demography: Multidisciplinary Approaches to Prehistoric Populations (ed. R. Paine). Center for Archaeological Investigations Occasional Papers No. 24. Carbondale, IL: Southern Illinois University Press, pp. 205–27.

Kohler, H.-P., F. C. Billari, and J. A. Ortega. 2002. “The emergence of lowest-low fertility in Europe during the 1990s”. Population and Development Review 28 (4): 641-680

Kelly, Robert L. 1995. The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington, D.C.: Smithsonian Institution Press.

Kermack, W. O., McKendrick, A. G. and McKinley, P. L. (1934), ”Death rates in Great Britain and Sweden: Some regularities and their significance”, The Lancet, March 1934.

Kremer, M. 1993, ”Population Growth and Technological Change: One Million B.C. to 1990”, Quarterly-Journalof-Economics 108, no 3.

Kuriyama, S., Ohmori, K., Miura, C., Suzuki, Y., Nakaya, N., Fujita, K., Sato, Y., Tsubono, Y., Tsuji, I., Fukao, A., Hisamichi, S., 2004, Body mass index and mortality in Japan: The Miyagi Cohort Study. Journal of Epidemiology 14, S33-S38.

Kuh, D. and Ben-Schlomo, Y. 1997. A Life Course Approach to Chronic Disease Epidemiology. Oxford: Oxford University Press.

Laara, E., Rantakallio, P., 1996, Body size and mortality in women: A 29 year follow up of 12,000 pregnant women in northern Finland. Journal of Epidemiology and Community Health 50, 408-414.

Lee, James Z. and Cameron Campbell. 1996. Fate and Fortune in Rural China. Cambridge:

Cambridge University Press.

Lee, James and Wang Feng. 1999. One Quarter of Humanity: Malthusian Mythologies and Chinese Realities, 1700–2000. Cambridge: Harvard University Press.

Lee, Ronald. 1987. “Population Dynamics of Humans and Other Animals.” Demography. November, 24:4, pp. 443–66.

Lee, Ronald. 1997. “Population Dynamics: Equilibrium,Disequilibrium, and Consequences of Fluctuations,” in Handbook of Population and Family Economics, Volume 1B. Mark Rosenzweig and Oded Stark, eds. Amsterdam: North Holland, pp. 1063–115.

Lee, Ronald D. 2003. "The Demographic Transition: Three Centuries of Fundamental Change." Journal of Economic Perspectives, 17 (Fall 2003), 167-190

Lee, Ronald and Michael Anderson. 2002. “Malthus in State Space: Macro Economic-Demographic Relations in English History.” Journal of Population Economics. 15:2, pp. 195–220.

Leon, D. A., Smith, G. D., Shipley, M., Strachan, D., 1995, Adult height and mortality in London: early-life, socioeconomic confounding, or shrinkage. Journal of Epidemiology and Community Health 49, 5-9.

Lesthaeghe, R. 1995. “The second demographic transition in Western countries: An interpretation”. In.: K. O. Mason and A.-M. Jensen (eds.) Gender and family change in industrialized countries. Oxford, Clarendon Press, pp. 17-62.

Livi-Bacci, M. 1986. Social-group forerunners of fertility control in Europe, in The Decline of Fertility in Europe (Coale, A.J. and Watkins, S.C., eds), pp. 182–200, Princeton University Press

Lutz, Wolfgang, Maria Rita Testa, Vegard Skirbekk, 2006. The "Low Fertility Trap" Hypothesis, Paper presented at the Population Association of America (PAA) 2006 Annual Meeting, March 30 - April 1, Los Angeles, California

Macunovich, D.J. 2000, Relative Cohort Size: Source of a Unifying Theory of Global Fertility Transition? Population and Development Review, Volume 26 Issue 2, June 2000

Malmberg, Bo & Lena Sommestad. (2000). Four Phases in the Demographic Transition, Implications for Economic and Social Development in Sweden, 1820-2000. Arbetsrapport/Institutet för Framtidsstudier;Working Paper 2000:6

Malthus, Thomas Robert. 1986/1798. “An Essay on the Principle of Population.” In Antony Flew (ed.) Thomas Malthus: An Essay on the Principle of Population. New York: Penguin Books.

Malthus, Thomas Robert 1986/1803. “An Essay on the Principle of Population.” In Wrigley, E. A. and David Souden. (eds.) The Works of Thomas Robert Malthus. Vol. II and Vol. III. London: William Pickering.

Mangel, Marc and Stephan B. Munch, 2005. A Life-History Perspective on Short- and Long-Term Consequences of Compensatory Growth, the American Naturalist, vol. 166, no. 6, December 2005

Marmot, M. G., Shipley, M. J., Rose, G., 1984, Inequalities in death--specific explanations of a general pattern? Lancet 1, 1003-1006.

Metcalfe, Neil B and Pat Monaghan, 2003. Growth versus lifespan: perspectives from evolutionary ecology, Experimental Gerontology Volume 38, Issue 9 , 1 September 2003, Pages 935-940

McGue, M, J. W. Vaupel, N. Holm, B. Harvald, 1993. Longevity is moderately heritable in a sample of Danish twins born 1870-1880. Journal of Gerontolology, 48, B237 (1993).

McKeown, Thomas, (1976) The modern rise of population, London: Edward Arnold.

McKeown, Thomas. (1979) The Role of Medicine: Dream, Mirage, or Nemesis? Princeton, NJ.: Princeton University Press.

Moormann, A. M., Sullivan, A. D., Rochford, R. A., Chensue, S. W., Bock, P. J., Nyirenda, T. & Meshnick, S. R. (1999) J. Infect. Dis. 180, 1987–1993.

Neugebauer,Richard, Hans Wijbrand Hoek, Ezra Susser, 1999. Prenatal Exposure to Wartime Famine and Development of Antisocial Personality Disorder in Early Adulthood, journal of the American Medical Association, Vol. 282 No. 5, August 4, 1999.

Notestein, F W (1945): "Population - the long view", in Schultx, T W (ed.): Food for the World. Chicago.

J. Oeppen, J. W. Vaupel, 2002, Broken limits to life expectancy, Science 296, 1029 (2002).

Olshansky SJ, Passaro D, Hershow R et al, 2005. A potential decline in life expectancy in the United States in the 21st century. [Special Report.] New England Journal of Medicine. 2005; 352:1138–1145.

Pelletier, François, 2004. HIV/AIDS and Adult Mortality: Methodological Aspects And Challanges For Africa, United Nations, Population Division, New York

Preston, Samuel H. 1996, American Longevity: Past Present, and Future, Syracuse University, Maxwell School of Citizenship and Public Affairs,Center for Policy Research Policy Brief No. 7/1996.

Preston, Sam. 1980. “Causes and Consequences of Mortality Declines in Less Developed Countries During the Twentieth Century,” in Population and Economic Change in Developing Countries. Richard Easterlin, ed. Chicago: University of Chicago Press, pp. 289–360.

Preston, S. H., Hill, M. E. and Drevenstedt, G. L. 1998. Childhood Conditions That Predict Survival to Advanced Ages Among African-Americans. Social Science and Medicine, 47:1231–46.

Preston, S. H. and van de Walle, E. (1978), ”Urban French history in the nineteenth century”, Population studies, vol. 32.

Ravelli ACJ, JHP van der Meulen , RPJ, et al. 1998, Glucose tolerance in adults after prenatal exposure to famine. Lancet 1998; 351:173–7

Ravelli ACJ, JHP van der Meulen, C Osmond, DJP Barker, OP Bleker. 1999. Obesity at the age of 50 y in men and women exposed to famine prenatally. Am J Clin Nutr; 70:811–6

Rissanen, A., Knekt, P., Heliovaara, M., Aromaa, A., Reunanen, A., Maatela, J., 1991, Weight and mortality in Finnish women. Journal of Clinical Epidemiology 44, 787- 795.

Riley, J.C. 2001. Rising Life Expectancy: A Global History, Cambridge Univ. Press, New York, 2001

Roff, D. A., 1992, The Evolution of Life Histories (Chapman and Hall, New York).

Rollo, C. D., 2002, Growth negatively impacts the life span of mammals. Evolution & Development 4, 55-61.

Rosebooma TJ, JHP van der Meulena, C Osmonde, DJP Barker, ACJ Ravellia, J M Schroeder-Tankab, G A van Montfransb, R P J Michelsc, O P Bleker, 2000. Coronary heart disease after prenatal exposure to the Dutch famine, 1944-45 Heart, 84:595-598

Sala i Martin, X, 2002. 15 Years of New Growth Economics: What Have We Learnt? in The Challenges of Economic Growth, Ed. Norman Loaiza, Central Bank of Chile, Santiago

Samaras, T. T., Elrick, H., Storms, L. H., 2003, Is height related to longevity? Life Sciences 72, 1781-1802.

Senior, Nassau. 1831. Two Lectures on Population. Delivered before the University of Oxford. Easter Term 1928. London: John Murray, Albemarle Street.

Shanley, D.P., and T.B.L. Kirkwood, 2000. Calorie restriction and aging: A life-history analysis. Evolution 54:740-750.

Smith, G. D., Hart, C., Upton, M., Hole, D., Gillis, C., Watt, G., Hawthorne, V., 2000, Height and risk of death among men and women: aetiological implications of associations with cardiorespiratory disease and cancer mortality. Journal of Epidemiology and Community Health 54, 97-103.

Stearns, S. C., 1992, The Evolution of Life Histories, Oxford University Press, Oxford.

Stein Z, Susser M, Saenger G, Morolla F. 1975. Famine and human development: the Dutch hunger winter of

1944–45. New York: Oxford University Press.

Schofield, R. S. (1984), ”Population growth in the century after 1750: the role of mortality decline”, in Bengtsson, T, Fridlizius, G, and Ohlsson, R., eds., Pre-industrial population change. Lund: Almqvist & Wiksell International, pp. 17-40.

Song, Y. M., Smith, G. D., Sung, J., 2003, Adult height and cause-specific mortality: a large prospective study of South Korean men. American Journal of Epidemiology 158, 479-85.

Susser ES, S.P. Lin (1992), Schizophrenia after prenatal exposure to the Dutch Hunger Winter of 1944-1945, Arch Gen Psychiatry. 1992 Dec;49(12):983-8.

Szreter, S., and E. Garrett 2000 Reproduction, compositional demography, and economic growth in early modern England. Population and Development Review 26:45-80.

Taeuber, I. B. (1945). "The Demographic Statistics of Southern and Eastern Asia." Journal of the American Statistical Association 40(229): 29-.37

Thompson, W. S. 1929. Population. American Journal of Sociology, vol. 34, No. 6, pp. 959-975.

Towner, M. 2001 Linking dispersal and resources in humans: Life history data from Oakham, Massachusetts

(1750-1850). Human Nature 12:321-350.

Towner, M. 1999 A dynamic model of human dispersal in a land-based economy. Behavioral Ecology and Sociobiology 46:82-94.

United Nations Population Division. 2002. Report Of The Expert Group Meeting On Completing the Fertility Transition. New York: UN Department of Economic and Social Affairs

Vaupel J. W. 2005. The biodemography of aging. In: LJ Waite (Ed.). Aging, health, and public policy: demographic and economic perspectives. New York: Population Council, 2005: 48-62

Vaupel JW, Baudisch A, Dölling M, Roach DA and Gampe J. 2004. The case for negative senescence. Theoretical Population Biology. 65(2004)4: 339-351.

Vaupel, J. W. 2003. Post-Darwinian longevity. In: J. K. Carey and S. Tuljapurkar (eds). Life Span. Evolutionary, Ecological and Demographic Perspectives. New York: Population Council, 2003:258-269, Population and Development Review, 29, Supplement

Voland, E, 2000 Contributions of family reconstitution studies to evolutionary reproductive ecology. Evolutionary Anthropology 9:134-146.

Voland, E., and R.I.M. Dunbar, 1997 The impact of social status and migration of female age at marriage in an historical population in north-west Germany. Journal of Biosocial Science 29:355-360.

Watkins, S.C. (1990) From local to national communities: the transformation of demographic regimes in western Europe, 1870–1960, Popul. Dev. Rev. 16, 241–272

Wienpahl, J., Ragland, D. R., Sidney, S., 1990, Body mass index and 15-year mortality in a cohort of black men and women. Journal of Clinical Epidemiology 43, 949-960.

Wilmoth J.R., L. J. Deegan, H. Lundström, and S. Horiuchi, 2000, Increase of Maximum Life-Span in Sweden, 1861-1999, Science 29 September 2000: Vol. 289. no. 5488, pp. 2366 - 2368

Wilson, Chris (2001) Implications of global demographic convergence for fertility theory. In Proceedings International Union for the Scientific Study of Population: working group on low fertility, Tokyo, Japan.

Wrigley, E. A. 1981. "Marriage, Fertility and Population Growth in Eighteenth-Century England," in R. B. Outhwaite, Marriage and Society: Studies in the Social History of Marriage.

Wrigley, E. A. and R. Schofield. 1983. “English Population History from Family Reconstitution: Summary Results, 1600-1799,” Population Studies, 37: 157-184.

Van de Kaa, Dirk. 1987. “Europe’s Second Demographic Transition.” Population Bulletin. 42:1, pp. 1–57.

Van der Zee, Henri A, 1998, The Hunger Winter: Occupied Holland 1944-1945, University of Nebraska Press

Yuan, J. M., Ross, R. K., Gao, Y. T., Yu, M. C., 1998, Body weight and mortality: a prospective evaluation in a cohort of middle-aged men in Shanghai, China. International Journal of Epidemiology 27, 824-832

The Second Stage

2008-02-13T12:45:00.000-08:00

As we have seen, for most of the second half of the twentieth century debates about population and population policy largely centered around the question of rapid population growth in the less developed world. Towards the end of the century, however, a quite different demographic phenomenon began to attract attention: aggregate fertility levels in the developed world that were inadequate for the long-run replacement of the existing population.

Analytically, the potential population policy issue raised by low fertility is not dis-similar to the problem inherent in rapid population growth: there is a disjunction between the sum total of individual reproductive decisions and a collective interest in some kind of long-run demographic equilibrium. Policy remedies, however, have been and remain much less clearcut.

In fact concern about below replacement fertility is not an entirely new phenomenon: the possibility that population might one day peak and start to decline could, for example, alreadt be discerned from a close inspection of fertility rates in Western European countries in the 1920s and 1930s, and in some specific cases, most notably France, even earlier. The birth deficit that began to attract attention in the pre WWII environment even had some communality with the more recent fertility decline in that it was associated with the start of a generalised postponment of first births (in this case possibly associated with economic and then war-related uncertainty). However the decades immediately following World War II were to see a substantial and generalised baby boom across the industrialised world (this time driven by a lowering in age at first birth) and preocupations waned with the same rapidity as the babies arrived.

The boom, however, was not to last, and by the 1970s, total fertility rates (TFRs) in most European countries and in the United States had once more turned downwards, and this time it seems that the change was a definitive one with falling in many countries decisively below replacement level. As a result fertility decline came back once more on government agendas as a policy issue. Subsequently fertility in the United States has stabilized at a level which is very close to the replacement rate (2.1), but in Europe and elsewhere fertility has continued to decline. By the beginning of the twenty-first century, the average TFR in Europe was 1.4. Such a level, if maintained indefinitely, will result in a population loss of one-third from one generation to another, that is population would fall by 30% roughly every 30 years. In some countries, notably in Southern, Central, and Eastern Europe, period fertility rates have fallen to 'lowest-low' levels without historical precedent. If this state of affairs continues, in the absence of large compensatory immigration it will lead not only lead to rapid population decline but also to very high proportions of the population in the older age categories.

Rising Median Ages

The decine in fertility has been accompanied by a continuous rise in life expectancy. Average worldwide human life expectancy reached 63 years in 1998 (CIA World Factbook 2004), with country values ranging from 37 in Sierra Leone and Zambia to 81 in Japan and San Marino. Indeed average life expectancy has increased in many developed countries in more or less linear fashion at a rate of nearly at 4 months a year over the last 160 years (Oeppen and Vaupel 2003, Riley 2001). The combined effect of these two processes - declining fertility and rising life expectancy - has been a steady rise in median ages. The median age of a population is the age at which 50 per cent of the population is older - and naturally 50 per cent is younger - than the age in question. Movements in median age can be considered a convenient rule of thumb measure of population ageing, although it is not an entirely unproblematic one (see later in this chapter and Sanderson and Scherbov, 2005).

According to the United Nations the 'global' median population age rose by a mere three years between 1950 and 2000, increasing from 23.6 to 26.4. (United Nations Population Division, 2002). The limited scale of this increase was largely the result of the fact that the most populous nations in the less-developed and developing world continued to have relatively high levels of fertility and thus remained extremely young.

(......something here on Wolfgang Lutz and probabilistic forecasts).

Over the coming 50 years, however, the global median age is expected to rise by nearly 10 years, being projected to reach 37 by 2050. What's more in 2050 13 currently developed countries are expected to have a median age of 50 or over, with Japan, Latvia and Slovenia (each with a anticipated median age of about 53), and the Czech Republic, Estonia, Italy and Spain (with a median ages of around 52) heading the list. In addition, three countries currently classified as developing (Armenia, the Republic of Korea and Singapore) will also be in the over 50 group. At the other end of the spectrum it is expected that Angola, Burkina Faso, Mali, Niger, Somalia, Uganda and Yemen will still young have comparatively young populations, with median ages likely to remain under 25. Clearly such projections are subject to high levels of uncertaintly, and in particular will depend on the pace of development in the countries concerned, a pace which will depend, in part of course, on the rate at which fertility itself falls.

In fact the entire planet is in the midst of a sea change in its age structure. The global population of over-60s, which is currently about half the size of the global 15-24 group, is rising sharply and is projected to surpass the 1 billion mark within two decades (and in so doing to become proportionately larger than the projected 15-24 age group). The over-80 population is expected to rise at an annual rate of 3.4% from 2000 to 2050, constituting an increase from 1% to 4% of the global population for this 'oldest old'group. Moreover, the process of population aging is itself accelerating. During the last 50 years, the global population of over 60s or over rose by just 350 million (to a total of 550 million), while in the next 50 years a 1.5 billion increase is projected for this group.

Since about 1970 the global overall dependency rate has being falling, due to the age-structural effects of falling fertility and declining levels of youth dependency. However, within this general trend old age dependency has been steadily rising, and from 2010 onwards there will be a steady upward trend in the overall global dependency rate as increase in the numbers of elderly outweighs the decline in the number of children.

This combination of these two trends - the systematic decline in fertility to below replacement level and the continuing increase in life expectancy which produces a steadily rising proportion of elderly dependent people - has been labelled by some the second demographic transition (Van de Kaa, 1994, 1987 Castles, 2003; Lochhead, 2000; Lesthaeghe and Moors, 2000). Others, on the other hand, continue to refer to societies exhibiting these characteristics as 'post transitional' (Bongaarts, 2001). Some have gone even further (Kohler, H.-P., F. C. Billari, and J. A. Ortega. 2002) and have started to talk of a third, or postponement, transition, thus treating the fertility phenomenon as a separate entity in and of itself. Kohler et al, in support of this view, argue that the long-term trend towards low and lowest-low fertility in Europe is related to three distinct transition processes:

(a) the (first) demographic transition which is characterised by parity-specific stopping behavior within marriage (ie less third, fourth, fifth children etc) and the growth of the nuclear family

(b) the second demographic transition which is characterised by sweeping value changes and in the rise of non-marital family forms

(c) finally, the third 'postponement' transition which is characterised by a steady shift in the timing of fertility towards a late childbearing regime.

The postponement transition is thus seen as a third step, one which follows-on from the control of marital fertility and the transformation in partnership attitudes and behavior.

Until recently relatively little attention had been given to the determinants and consequences of fertility in 'below-replacement' fertility societies. Traditional theories of the demographic transition in fact had comparatively little to say about the level at which fertility would stabilize at the end of the envisioned transition. It was normally assumed - or implied - that replacement fertility of about 2.1 births per woman would prevail in the long run (Demeny 1997; Caldwell 1982). The undelying theoretical justification for this assumption would seem to have been that the transition would be from one homeostatic equilibrium to another, in which case a stable and sustainable fertilty level would be one of the characteristics of the ultimate end state.

Surprsingly or unsurprisingly this homeostatic assumption has continued to influence thinking and is, for example, to be incorporated in the population projections of global institutions like the UN and of the World Bank (medium variants). Such projections normally anticipate that fertility will stabilise at around the 1.9 level in the long run (the justification for this seems to be a calculation based on a termination of the tempo effect, but we have really no good reason to assume that the downward quantum process will itself come to an end). One of the underlying reasons for the existence of such a parity specific interpretation of what a homeostatic process might look like may well be a a result of failing to appreciate just what the long run impact of a steady and sustained increase in life expectancy might be. If fertility and longevity are interconnected in the way Life History Theory suggests that they might be, then there coulb be good reasons for anticipating that fertility will slowly trickle downwards just as longevity trickels upwards (Kaplan et al, 2003).

Stylised Facts of the "Second Stage"

Those who argue that the most recent fertility regime which is to be found in the developed world constitutes a second stage, or second transition, normally point to a number of stylised facts which serve to characterise it. (A good general introduction to the 'second stage' issues can be found in Sleebos, 2003, which proved expecially useful to the author in preparing this section).

1/. There has been a generalised decline in fertility rates - starting in the 1970s - in most OECD countries.

2/. There have been significant differences in the way this decline has been experienced in individual OECD countries (the Nordic and English-speaking countries, on the one hand, where the decline in fertility started earlier but later stabilised or even - in particular in the case of the US - reversed, and the Southern European countries, where the decline started later but then proceeded much faster). One consequence of this differential pace of decline is that some of the countries who were at the top of the OECD fertility league just a few decades ago are now close to the bottom, whilst others (the United States, France) which had been experiencing decline have either stabilised or even regained a little ground.

3/. There has been a general increase in the mean age of women at first childbirth across the OECD countries, but as with the fertility decline itself there have also been differences across countries in the extent of fertility-recuperation at higher ages, and differences in the distribution of births according to mothers’ age (with persistently high teenage-births characterising the fertility regime in some countries, Sobotka, 2004).

4/. There have been systematic differences across the OECD in fertility rates among women with different characteristics (with higher fertility being found among non-working women, women who work part-time, women living in married couples and women coming from either ethnic minority groups or arriving in the course of migratory movements). At the same time there has been considerable variation across countries in the size of these differences.

5/ A continuing gap has been found in most OECD countries between realised and desired fertility, with the latter (at least until very recently) tending to remain clustered around the "two-child" norm. (Evidence is growing that this may be changing, at least in some countries, see Goldstein, Lutz and Testa, 2003).

6/ There has been a tendency for fertility rates to be lower in OECD countries with:

a) lower employment and educational attainment of women

b) lower frequency of divorce and of out-of-wedlock births (or simply with more 'familiaristic' welfare traditions).

c) higher obstacles for young people in making the transition from school to the establishment of an independent home.

d) higher income of elderly people (although this finding is not as strong as the previous three). (Reference, please???).

Those wishing to hold to a strong version of 'second transition' (or value transition) theory do not cease to point to the fact that the pattern of association of some of these above-mentioned variables with fertility rates has reversed its sign relative to the relation which was the case only a few decades ago. This fits in well with their argument since amongst second transition theorists there is one common and recurring theme: that recent fertility trends have been accompanied by notable changes in attitudes and behaviour regarding sexuality, marriage, and family and household formation. Such changes include higher levels of cohabitation, pre- and extramarital childbearing, widespread use of abortion and contraception, rising levels of childlessness and divorce, as well as delays in age at first marriage and first birth. Ironically it was previously the nuclear family which was seen to be the bedrock of a stable (self reproducing) society, whilst today it is childbirth out of wedlock which is seen as the key ingredient of stable fertility.

Ideal Family Size

More recently, work on below-replacement fertility has focused attention on how social and economic development and changing ideas and values (post-modernity) serve to influence the desired number of children women wish to have (van de Kaa 1998). Earlier versions of these theories often assumed - whether implicitly or explicitly - that couples are able to implement their preferences without too much difficulty and hence the conclusion was drawn that observed long run fertility would not differ greatly from average desired family size. Whilst it is clear that declining desired family size is indeed one of the principal forces driving fertility transitions, many demographers are now becoming aware that levels of fertility may often deviate substantially from stated preferences, and this has lead to increased focus and attention being directed towards the presence of what may be considered to be 'inhibiting factors' (Lutz et al, 2006).

Fertility preferences are themselves not without ambiguity, and several concepts have been used in attempts to identify and measure them: ideal number of children, desired number of children, expected number of children (in addition to those already born), intended number of children etc etc.

The main objections to this type of procedure usually relate to the use of the concept of 'ideal family size' itself. Some authors have criticised the abstract character (Toulemon 2001) of an indicator which is aimed at measuring poorly-defined concepts such as reproductive goals, whilst others question the the validity of demographically non-specific questions which do not refer, say, to concrete factors like the already experienced childbirth (van de Kaa 2001). Yet others point to the lack of sensitivity of the concept to fluctuations in actual realised fertility (Livi Bacci and De Santis 2001).

The whole procedure of examining future childbearing preferences seems to be intrinsically fraught with difficulty. With reference to ideal family size, for example, the question might be asked as to the kind of validity which can be assigned to the reliability of the various measures used (ideal, desired, intended or expected fertility). How, for example, should such questions be worded, how does one overcome the difficulties of comparing responses across countries, what significance should be attributed to nonresponse cases and 'don’t know' answers, or even, just how useful at the end of the day are these kind of questions as predictors of future actual childbearing? (Sobotka, 2004)

All of this having been said, it is not difficult to understand the reason why the persistent and clear discrepancy between actual reproductive behaviour and declared fertility preferences, regardless of the indicator used, has continued to exert a strong attraction for those who seek to understand the low levels of realised fertility which now characterise the majority of countries in the developed world (Van Peer 2002).

The gap between intentions and outcomes is typical of the OECD world where desired family size is still normally around the two children mark while achieved fertility has almost universally fallen below (or well-below) replacement. This divergence between desired and achieved fertility is of more than passing theoretical interest sinces it raises the important question of whether the low fertility observed in the developed world is depressed as a result of transient tempo factors, or whether a more lasting process may be at work.

If the current state of affairs were merely a transient one, then fertility could be expected to rise to a level closer to the preferred level in the not too distant future (this is the view, for example, of a respected demographer like Masimo Livi-Bacci), and in the event of this being the case concern over the possible undesirable economic and social implications of prolonged low-low fertility in developed societies would be largely misplaced. On the other hand, were below replacement fertility to form part of an ongoing process, then the long term economic and social consequences would be far from insignificant.

For this reason the most recent Eurobarometer survey (European Commission, 2001) has been closely scrutinied. The survey revealed that while in most EU countries women of all ages still held to family size ideals which were above replacement level, in Austria and Germany the national averages had fallen significantly below this level (Goldstein et al, 2003). Indeed younger cohorts in Austria and Germany reported even lower ideal family sizes than the average, with results in the region of a desired family size of 1.7 children. It is not without interest to note here that Austria and Germany were among the first countries to experience declines in period fertility to well below replacement levels, and this result does at least raise the question as to whether the changing reality of actual family size may not, finally, be beginning to influence the ideals of subsequent generations. More recent evidence coming from the EU funded DIALOG study (which interviewed 30,000 people in 14 European countries on their attitudes and opinions concerning family numbers, fertility behaviour and demographic change) found that in the main the two child 'ideal' was still common across Europe, but that below replacement ideals existed in Germany, Italy, Austria and Belgium and the Czech Republic (The desired number of children per women in Germany was found to be (1,75), in Italy (1,92), in Austria (1,84), in the Czech republic (1,97) and... in the Flanders region of Belgium(1,86), Dialog Project, 2006).

The emergence of below-replacement ideal fertility, were it to persist,could mark a new stage in the evolution of below-replacement fertility regimes. As Bongaarts (2001:276) writes: "whether desired family size remains at or drops below two is the most crucial issue determining post-transitional fertility."

Despite all the above mentioned measurement difficulties, changes in fertility preferences still play a causal role in most theories of fertility decline (Lesthaeghe and Surkyn 1988, Van de Kaa 2001). Recent research on the dynamics of behavioural diffusion establishes a strong relation between desired and achieved fertility (National Research Council 2001) and theoretical models such as that built by Kohler (2001) assume a strong connection between the achieved fertility of others and one’s own desired fertility.

For Goldstein et al (2003) the most plausible explanation for the German and Austrian findings is that sub-replacement fertility ideals in those countries have emerged as a natural consequence of a history of low-fertility. Young cohorts in Germany and Austria have witnessed below-replacement fertility for their entire lives, not just the last few years. The influences of a cohort’s experience on their present fertility ideals are probably numerous. Young Austrians and Germans are not only more likely than other Europeans to have grown up in smaller families; they are more and more likely to have had friends, classmates, and cousins in smaller families as well.

So it is worth considering the possibility here that low family-size ideals may create a momentum of their own, making it difficult for would-be pro-natalist policymakers to raise fertility levels in the future. Whatsmore, if the generational-lag interpretation of fertility preferences in Germany and Austria is confirmed, we may be about to see the emergence of falling family-size ideals in other low-fertility countries, like Italy and Spain, in the not too distant future.

The Historical Record

Fertility in the developed world reached its post-World War II peak of around 2.8 births per woman during the baby boom in the late 1950s. Subsequently steep declines in fertility in the 1960s and 1970s left TFRs well below replacement, and, as has been repeatedly stressed here, in some cases extraordinarily so.

In the late 1950s regional fertility levels in the developed world ranged from a high of 3.7 births per woman in North America to a low of 2.1 in Japan, but they later converged, and by 1980 the global average was in the region of 1.8 births per woman. Since 1980 fertility levels have diverged again, with US fertility rising to a continuing 2.0 births per woman while reproductive levels in Japan and Europe have continued to remain low. In fact the United States is a major demographic 'outlier' in this whole 'transition process', since the fertility lows of the late 70s have been reversed under the impact of an unprecedented immigration shock on the one hand, while the continuing presence of high rates of adolescent pregnancy, on the other, have meant that the mean first childbirth age is still remarkably low by first world standards.

This being said, country specific variations in fertility within regions can be substantial. For example, in the European context fertility is lowest in the south, where sharp declines have occurred since 1975, and in the east, which has seen a dramatic fertility reduction post 1989. At the present time Italy, Spain and Japan - with recorded TFRs in the 1.2 to 1.3 range and high and increasing levels of life expectancy - are competing for the rather dubious status of being the world's most rapidly ageing nation.

In contrast fertility in Northern Europe, which averages nearer the 1.8 TFR mark, is significantly higher, and has shown signs of being relatively stable. Perhaps the most marked European contrast is the one to be found between France and Germany: the German TFR is close to 1.3 while the French equivalent is around 1.9. This shows that talk of a European 'average fertility rate' may, in fact, be deeply misleading.

Steep declines since 1960 have also left fertility below replacement in Thailand (TFR 1.7), and well below replacement in Singapore, Korea, Hong Kong and Taiwan (TFRs in the 1.2 to 1.3 range). The application of a rigid neo-Malthusian policy of one child per family in mainland China has also lead to a rapid decline in fertility and a TFR which is well-below replacement level (the Chinese data are notoriously unreliable it should be noted: see Lutz et al, 2005).

Homeostatic or Not?

One of the key issues which has emerged in the debate which surrounds this 'second transition' fertility shift is whether the process of sustaining fertility at or around population replacement level is a homeostatic one or not. (Wolfgang Lutz takes a rather jaundiced view on this front. Indeed he is far from accepting that current levels of lowest low fertility are even themselves a 'floor'. The only justification he claims to have found for a return to replacement fertility in what is by now an extremely voluminous literature, is the suggestion that "in the end governments will do something about it).

Perhaps the best-known exponent of the view that sustaining near replacement population constitutes a homestatic process is the Italian demographer Masssimo Livi-Bacci (Livvi-Baci 2001). Livi-Bacci uses expressions like the current potential for population maintenance being 'badly hampered' in a way which he sees as being similar to the plight of medieval fertility under the impact of the plague, or 19th century Irish fertility during the Great Famine, or - in early modern times - to the effects of WWI. He is at pains to point out that on each previous occasion population growth eventually recovered, so why, he asks, should it not recover now from the current fertility shock?

Why indeed? Well, in the first place, as David Weil asks (Weil, 2001), is it really realistic to expect fertility rates to return to a level consistent with population replacement in the mature industrial societies? What justification is there for this assumption? Before jumping to any such a conclusion, shouldn't we ask ourselves a little about the source of these homeostatic tendencies in historical populations that Livi-Bacci claims to be able to discern? Here, of course, it is impossible not to invoke the name of that patron saint of all homeostatic population models - and perhaps the most dismal of all dismal economists - the very Reverend Thomas Malthus.

The 24 trillion dollar question is whether the weakly valid homeostatic process which characterised the pre-modern Malthusian regime still operates today? Weil suggests, and I agree, that we have little reason to think that it does. Firstly, the economic and demographic mechanisms which underlay the Malthusian model have been altered radically. In particular up until the end of the nineteenth century, it was possible to argue that higher incomes in and of themselves would lead to higher population growth. In the post-industrial societies of today, however, there now seems to be little evidence for any strong relation between population growth and levels of per capita income, indeed quite the contrary seems to be the case. At the same time, in the developing world the Malthusian relation seems to have inverted, so that higher income now increasingly leads to lower population growth. As one United Nations population conference famously concluded: "Development is the best contraceptive."

The second reason that the lessons of the Malthusian regime are probably not applicable today is that the nature of the shocks that impinge on population dynamics is very different, indeed, as is argued here, we may not even be dealing here with a shock at all, but rather with one continuing 'demographic transition' which spans the whole period which starts with the origins of the fertility decline at the time of the indutrial revolution itself.

Historically the sources of demographic shocks were things like plagues, wars, the discovery of new lands,improvements in agricultural productivity (like the arrival of the potato in Ireland) etc. But the 'shock' that underlies the current seismic shift resembles none of these phenomena. Rather, it is one continuous and seamless behavioural adaptation: an adaptation whose net consequence is that people simply want to have less children, and less, and less.....

Why this should be the case is a fascinating question, and one which we are systemtically attempting to address here, but it is not one that is likely to be answered simply by limiting our field of vision to the historical record in a search for ealier equivalents.

Once we depart from the Malthusian world, the homeostatic tendencies that Livi-Bacci argues for are simply no longer present. As David Weil argues there is nothing natural or 'god given' about zero population growth, and, thus, no particular reason why we should expect fertility in any country to be at or near the replacement level.

The Low Fertility Trap

Building on this insight, and in sharp contrast to the more 'optimistic' views of a Livi Bacci, the Austrian demographer Wolfgang Lutz has gone on to raise the concern that those societies experiencing lowest-low fertility may now be caught in some kind of low fertility trap (Goldstein, et al 2003, Lutz et al 2005, Lutz et al 2006)

Lutz's starting point is the question, just how low can fertility go? As he says “We lack any good theory about the future path of fertility. In fact we cannot say anything specific about a possible lower bound” (personal communication with author). Given this level of uncertainty the question logically needs to be asked, so just what is the likely future fertility path of those societies currently experiencing lowest-low fertility? Now we know that these fertility levels are produced by the combined impact of two associated but distinct processes - a birth displacement one and a reduced cohort completed quantum (or parity) one. The issue is complex since the posponement process can itself run for many decades (and yet again, we have no good reason for assuming a final end-point for the upward displacement in first birth ages, life history rescaling and assisted fertility processes may drive the upper boundary above levels which we can currently contemplate). The question of just what kind of 'fertility recovery' can be contemplated or expected in the lowest-low fertility societies.

Now, in this context, Lutz defines the basic idea lying behind his hypothesis as being the following: once fertility falls below a certain level and stays there for sufficient time this can produce a self-reinforcing demographic regime change that is difficult or impossible to reverse. He then proceeds to base the 'low fertility trap' hypothesis on the operation of three distinct but interacting mechanisms:

1/. A population momentum component. The delay in childbirth produces very low fertility rates which last for decades, during this time there is pyramid-base shrinkage, and new generations arrive composed of much smaller cohorts. This systematically produces less and less children.

2/ An ideas propagation mechanism. This works via the idea of 'ideal family size': young people are increasingly socialized in an environment with few children, and this may result in a lower 'ideal family size' in the subsequent generation, and so on. Recent (2001) Eurobarometer readings from Germany and Austria indicate that young people may now, on average, have a below replacement ideal of family size (slide on page 8). Since preferences and expectations are important here, this can only lead fertility downwards.

3/ A negative economic feedback process due to cohort and other effects (the Easterlin thesis). Lutz, Skirbekk and Testa develop a flow chart model (page 7 of this presentation in the Adobe Acrobat) which attempts to describe this process.

Lutz's idea originates from:

(a) the observation that countries which have fallen below fertility his critical level of 1.5 TFR have generally not subsequently 'recovered' in the sense of returning to a level above 1.5. In particular the germ of the idea here comes from the work of the Australian demographer Peter MacDonald (see this paper) who was the first to note that 1.5 TFR seems to mark something of a watershed.

(b) an awareness that whilst most projections and policy were being set by the assumption that there was a likely 'homeostatic' return to near replacement fertility, there is no rigourous theoretical justification for this assumption.

As Lutz et al observe:

"Virtually all population projections for low fertility countries assume end of fertility decline at current cohort level (Eurostat) or increase (UN), while at the same time continued increases of life expectancy are assumed. To be honest: we have no good theory with predictive power. Some “soft” arguments: end of postponement, children make happy, governments will eventually do “something”. But at the same time the basic forces that brought down fertility continue to work, possibly even stronger (value change, globalization,youth unemployment)".

(c) the further observation that government policy has non-linear consequences in this area. This view differs, for example, from that of the Canadian sociologist Anne Gauthier who argues straight and simple that: Public policies have an undeniable effect on families. On the other hand it is close to the work of social network theorists who examine propagation mechansisms for ideas and values across societies. In the words of Ronald Rindfuss and his co-workers:

“Changes in attitudes likely create a feedback mechanism, influencing behavior; and changes in behavior likely create a feedback mechanism influencing attitudes.” (Rindfuss et al. 2004, p. 855)

In the terms of Lutz et al: once the number of children (siblings, friends, children seen in other families, media) experienced during the process of socialization falls below a certain level, their own ideal family size may become lower which in course may result in further declining actual family size and still lower ideals in the subsequent generation.

The idea of negative demographic momentum is closely associated with the other key contribution Lutz has made to our understanding of the "second demographic transition": his idea of a 'birth deficit'. This deficit arises due to the the continuing presence of a fertility tempo effect, wherby the increase in the mean age of childbearing results in a lasting loss of births, and these 'missing' births cause structural damage to the age pyramid.

In the main Lutz bases his economic feedback mechanism on the cohort impact theory of Richard Easterlin and his associated 'relative income hypothesis. According to Easterlin changing cohort size produces either a crowding-out (the baby boom) or a crowding-in (declining fertility) phenomenon. The hypothesis posits that, other things being constant, the economic and social fortunes of a cohort (those born in a given year) tend to vary inversely with the relative size of that cohort, which is itself approximated by the crude birth rate in the period surrounding the cohort's birth. The cohort mechanisms operate mainly through three main social institutions – the family, school and labour market. Diane Macunovich has a good summary of Easterlins ideas and their application to fertility changes in Relative Cohort Size, Source of A Unifying Theory of the Global Fertility Transition.

The operation of this general 'crowding mechanism' means that large birth cohorts face adverse economic and social conditions, higher unemployment, and lower than expected wages, outcomes which are significantly at odds with their material aspirations. As a result, they postpone family formation and have fewer children. This line of research now represents a long-standing tradition in the United States, where an ongoing body of work (Easterlin 1976, 1978, 1987, Macunovich 1998a,1998b, 2000, 2002, Bloom, Freeman, and Korenman, 1987, Korenman and Neumark, 2000) has posited the idea that the relative size of young cohorts entering the labour market has far-reaching implications for wages, inflation, unemployment rates, etc, as well as for a variety of cohort impacting factors like living standards and family behaviour. It is also now being applied in studies of the 'greying' phenomenon in the United States as the large 'boom generation' steadily approaches retirement age. .

On the other hand, the crowding-in syndrome would mean that the reduced cohorts which follow the fertility decline should find work more easy to obtain, and salaries relatively higher. This should lead to rising income expectations, which may be more difficult to sustain as the fiscal burden weighs down on younger generations with the consequence that they continually postpone starting families.

This latter eventuality seems to have relatively little empirical evidence to date to back it up (except, perhaps, very recently in Japan) so should really be treated with some caution.

Macunovich takes the theory and tries to use it to develop a general theory of the whole demographic transition from cohort effects, and I feel that at this level the argument is not convincing. The cohort dimension is however very evident in the US baby-boom phenomenon, and the subsequent fertility reaction, and indeed this is having the consequence that population ageing is being seen very much as a cohort phenomenon in the United States, but this experience is hard to generalise.

Lutz et al do, however, offer another suggestive direction for analysis: low fertiliy leads to the acceleration of societal ageing, this produces cuts in welfare and pension benefits, generates a general pessimism about the future and lowers expectations about future income. The general pessimism, coupled with anticipations of increased life expectancy, can produce increased saving for the future, and this of course can produce a drag on current consumption. The drag on consumption produces a lethargic level of economic growth, and this induces young people to delay having children in order to attempt to maintain current income. This economic chain reaction, especially in the light of what is actually happening in Germany and Japan, does seem to be one of the possible mechanisms through which the trap - should it in fact exist - might operate.

However, at the end of the day, as Wolfgang Lutz himself emphasises, what we need is more evidence. Perhaps we will find some in the next Eurobarometer survey on family intentions (due June 2006).

Lutz and Skirbekk (2004) have argued that 'policies aimed at creating the conditions that allow women to have their children at an earlier age, or at least not being driven into further delays, could turn out to be win-win strategies, combining individual health concerns with public demographic concerns.' Clearly such policies, if successful, would invert the process and recover some of the missing births as a 'reverse tempo' process results in a - one off - gain in births. Lutz points out that this possibility has encountered a certain resonance in the recently declared policy aims of the Family Federation of Finland, which now has as one of its objectives an increase in total realised fertility from 1.8 to 1.9, and this (where they to be succesful) would be achieved via a reduction in the average age at first birth from 28 to 26 years (Söderling 2005).

In support of this suggestion Lutz and Skirbekk estimate that merely terminating the process of fertility postponement in the Finnish case would bring the period TFR close to the established target of 1.9 (1.86), while any rejuvenation of fertility would push total fertility significantly above this level. However, we strongly need to bear in mind that in this area having the 'best of intentions' is simply not enough. Positing an objective is one thing, and achieving it another. If the demographic transition process has an inbuilt dynamic towards steadily higher ages at first birth and steadily lower levels of fertility, then there may be little that policy can do in the longer term to reverse this. We may be simply doing a kind of modern version of the labour of Sisyphus. What such policies may however be able to achieve is a softening of the overly dramatic (and demographically speaking structurally harmful) decline which we are currently seeing in country after country - and a like initiative from the Bavarian Parliament in Germany to compress the secondary school cycle by one year could be seen in a similar light. In any event making would-be mothers aware that they should have desired children earlier, and facilitating this approximation of wishes to reality, is surely a sound idea in and of itself.

The Postponement In Childbearing

2008-02-12T12:47:00.000-08:00

All the indications point to the possibility that sometime within the next decade the majority of the world’s population will find itself living in regions of the globe which have either near-replacement or below-replacement levels of fertility. What were previously thought of as distinct fertility regimes, with a strong distinction being drawn between fertility in "developed" and "developing" societies, are increasingly coming to be seen as forming part one single fertility phenomenon (Bongaarts and Bulatao 2000; Lutz et al. 2001; Wilson 2001).

Several aspects of this convergence towards low fertility are particularly striking.

First, the spread of below replacement fertility to formerly high fertility countries has occurred at a remarkably rapid pace and this has had the not insignificant consequence that the global convergence of fertility levels has notably preceded the convergence of many other social and economic indicators.

In Southern Europe (Spain, Italy, Greece and Portugal), for example, fertility started to fall much later than it did in Northern or Western Europe, but when it did finally start to fall it did so at a much faster rate than those who bagan the demographic transition earlier (Bagavos and Martin, 2001). By the time we get to the mid 1990s total fertility rates in Spain, Italy and Greece had already fallen so far that they were registering 'lowest-low' levels of below 1.3 TFRs. The recent pace of fertility decline has also been particularly rapid in the former Communist countries, such as Slovenia, Croatia, the Slovak Republic, the Czech Republic and Poland, all of whom saw their total fertility rate decline from around its near-replacement level of the mid-1980s to around 1.3 recently. Indeed, in former East Germany, the total fertility rate was cut in half between 1990 and 1992, reaching an unprecendented level of 0.9 in the immediate aftermath of reunification. (Sobotka, 2002)

Secondly, earlier ideas that fertility levels would naturally stabilize close to replacement level have been rudely shattered (see the last section). Far from convergence at replacement level, what we have actually seen has been the arrival of extremely low levels of fertility in some countries. Starting with the example of Spain and Italy in the early 1990s, the phenomenon of 'lowest-low' fertility - defined as total fertility rate levels below 1.3 (Kohler, Billari, and Ortega, 2002) - has spread and spread to such an extent that it now encompases some 13 lowest-low fertility countries in Southern, Central and Eastern Europe, as well as Germany, Armenia, Singapore, South Korea, Hong Kong, Taiwan and Japan. What is more, the pattern of evolution of fertility rates in some developing countries is giving rise to some concern that the number of these 'lowest-low' fertility countries may well have a tendency to increase in the near furute as first birth ages rise rapidly in what may be considered to be generationally short periods of time (ie in timespans of fractions of a generation).

Thirdly, fertility trends within the developed world continue to reveal surprisingly diverse patterns, with European fertility levels ranging from those found in the lowest-low fertility countries to the seemingly more benign TFR levels to be found in France (1.9), Denmark (1.8), or in the UK and Sweden (1.7). At the same time Japan and the Asian tigers now have versions of lowest-low fertility, while the United States has 'revived' and now sustains TFR levels within striking distance of the 2.1 population replacement rate.

Fourthly, we have seen the steady and continuous evolution of what Lance Pritchett in another context called 'big-time divergence' between the core global trend in fertility and life expectancy and the performance of a small group of demographic 'outliers'. In particular the 2005 United Nations Human Development report identifies 18 countries as suffering from either stubbornly high fertility rates, or from rising mortality rates, or from a combination of the two (United Nations, 2005). As the UN notes, male life expectancy at birth in the Russian Federation - at 60 years - is now the lowest for any industrial country, and this is a situation which compares starkly with that of Russia's West European neighbours, where average life expectancy is now around 79. Since the early 1990s there has been a marked increase in male mortality over and above the historical trend in the Russian Federation, with the number of additional deaths during 1992–2001 being estimated at between 2.5 and 3 million. In the absence of war, famines or health epidemics there is no recent historical precedent for the scale of this loss of life.

A look at the immediate causes of death does at least provide us with part of the explanation for what is happening in Russia. Firstly it is clear that there is a high presence of cardiovascular disease among adult males, something which in part at least reflects dietary and lifestyle influences. But alongside this 'First World' epidemic, the Russian Federation is also increasingly affected by infectious disease problems, with tuberculosis and HIV/AIDS in the forefront. Homicide and suicide rates are also compartively high by industrialised country standards, and both of these have seen marked increases since the early 1990s. In both cases the these indicators are closely associated with the overconsumption of alcohol.

All of this raises the question of whether Russia (with a relatively high median age of 38.15 and a relatively low TFR of 1.4) may not now be caught in another kind of demographic trap. With a sizeable disparity between male and female life expectancy, and compartively few children, it is hard to foresee the future of a society with such a distressingly warped population structure. Unfortunately we may face a future where the majority of Russian men grow neither old nor rich.

Such exceptions apart, viewed globally, the life-expectancy gap is still closing, although this needs the important qualification mentioned above that since the early 1990s the generalised long-run trend towards convergence has been significantly affected by regional divergences linked to the rise of HIV/AIDS and a number of other health issues.

And those with the lowest levels of life expectency may normally be expected to be those who experience the most rapid increases in life expectancy as they move towards the lifespan frontier after they begin their fertility decline. Between 1960 and 2005, life expectancy increased by 16 years in developing countries and by 6 years in developed countries. However due to the 'outlier' phenomenon mentioned above, the aggregate pace of increase has slowed, with the gap only closing by two years between 1980 with 85% of the two year convergence taking place in the 1980s. Indeed since that time, convergence has virually ground to a halt, and the overall spread remains very large, with the diference between a low-income country and a high-income one still being 19 some years, with someone, for example, born today in Burkina Faso expecting to live 35 years less than someone born in Japan, while someone born in India may expect to live 14 years less than someone born in the United States.

HDI reversals play an important role in the relative standing of a number of countries. In 2004 an estimated 3 million people died from the virus, and another 5 million became infected. Almost all of these deaths were in the developing world, with 70% of them in Africa. Some 38 million people are now infected with HIV—25 million of them in Sub-Saharan Africa.

In Sub-Saharan Africa the lethal interaction of economic stagnation, slow progress in education and the spread of HIV/AIDS has produced a free-fall in the Human Development Index ranking of a number of countries. Southern Africa accounts for some of the steepest declines of all - a fall of 35 places for South Africa, 23 places for Zimbabwe and 21 places for Botswana. Of the countries of the former Soviet Union the biggest declines were in Tajikistan, which fell 21 places; Ukraine, 17 places; and the Russian Federation, 15 places.

Sub-Saharan Africa is the region that explains the greater part of the slowdown in progress towards greater global equality in life expectancy. Twenty years ago somebody born in Sub-Saharan Africa could expect to live 24 years less than a person born in a developed country, and the gap was shrinking. Today, the gap is back at 33 years and still climbing.

Statistics alone cannot capture the full scale of suffering associated with the impact of HIV/AIDS, but they can provide an insight into the scale of the demographic shock inflicted on the worst-affected countries. On current indicators a child born in Zambia today has less chance of surviving past age 30 than a child born in the United Kingdom in 1840.

So at one pole of the globe we still have societies with continuing high mortality, and continuing high fertility, while at the other we find the opposite tendency with rectangular mortality low mortality and low fertility increasingly representing the norm.

A catchy way of putting all this would be to say that we have not one population problem but two, except that, as we shall see, the existence of a high-life-expectancy low-fertility regime isn't really a problem in and of itself (even if the adjustment process to one might be) all of this really forms part of one continuous seamless process.

Lowest-Low Fertility

One serious attempt at an explanation of lowest-low fertility which has emerged in Europe and elsewhere has been based on a combination of four distinct demographic and behavioral factors (Kohler, H.-P., F. C. Billari, and J. A. Ortega. 2002).

1/ The extremely low fertility readings are in many ways a statistical construct. As Bongaarts and Feeny have argued (Bongaarts and Feeney, 1998) conventionally used period fertility measures (like TFRs) show readings which are the product of a combination of both birth postponment and changes in the parity composition of the childbearing population. This means that the reading given on the derived period fertility indicators is, in all probability, significantly below the likely actual final level of achieved cohort fertility.

The derived TFR is, in fact, subject to both what are called 'tempo' and 'compositional' influences. In simple terms the 'tempo' influence arises from the postponement of childbirth and the result is a statistical indicator which be very misleading if what we are really interested in is the period-based final-outcome TFR. Clearly subsequent cohort-based achieved-data offers a more reliable indicator (and the end result is invariably significantly higher), but equally clearly such data only becomes available as the relevant cohorts reach the end of their fertility window, that is it only becomes available after a considerable lag, and this in itself complicates an already complicated policy decision process.

The other element - the so-called compositional effects - refer to movements in final achieved fertility levels, and their composition in terms of what are know as birth 'parities' (that is to say how many women are childless, and how many have four or five children, etc, etc).

The above-mentioned 'tempo distortion' clearly forms part of the lowest-low fertility picture, although its degree of importance is subject to ongoing debate since the extent of the subsequent fertility 'recovery' has been contested, and some demographers (most notably Wolfgang Lutz) have been at pains to explain that whatever the final achieved cohort outcomes the tempo effect irreversibly changes the shape of the pyramid (unless there is a later 'compensatory' advancement of birth ages), and the size of the reproductive base diminishes in any given society (ie you get generational-shrink).

Thus, while a focus on period TFRs offers an easy, and extremely user-friendly, classification of lowest-low fertility, unadjusted TFR readings need to be handled with care due to the inherent problems they conceal.The difficulty involved in interpreting these distortions has received a good deal of attention in recent times due to importance of delay in low fertility societies (e.g., Bongaarts and Feeney 1998, Kohler and Ortega 2002, Kohler and Philipov 2001).

2/. Ongoing economic and social changes make the postponement of fertility a rational response for most individuals. The higher educational demands of the labour market are often mentioned in this regard, and particularly the presence of an increasingly educated female population. Insecurity is another often mentioned topic, with labour market reforms, the widespread use of temporary employment contracts and the globalisation of labour and product markets being frequently mentioned in this context (Blossfeld, H.-P. 2005).

3/. Social-interaction processes which impact on the timing of fertility mean that the overall demographic and economic consequences of a changing socio-economic environment is much greater than the sum of the direct individual responses would lead one to expect (or put another way these are non-linear processes). Indeed quite modest socioeconomic changes appear to lie behind the rapid and persistent postponement phenomenon which we see reflected in the recent generalised onset of low and lowest-low fertility.

4/. Institutional settings in Southern, Central and Eastern Europe, as well in many newly developed countries (Taiwan, Singapore, Japan, South Korea), may well have favored an overall low quantum of fertility (McDonald, 200, 2002)

Postponement-quantum interactions have amplified the consequences of these institutional settings, and as a resuly have caused particularly large reductions in completed fertility in lowest-low fertility countries.

It is important to bear in mind that it has also often been argued that lowest-low fertility may not lead to particularly low completed cohort fertility if what is involved is primarily a temporal phenomenon. In this case, fertility eventually could eventually ‘recover’ through increased childbirth at the older ages (Frejka and Calot 2001a,b; Lesthaeghe 2001; Lesthaeghe and Willems 1999).

This debate is important since, as we have noted, in recent years a general and progressive delay in the age of the first childbirth has been observed throughout the OECD world and in particular in every single European Union country. The percentage of births to mothers aged thirty or over now exceeds 40% of total births in a number of countries, including Sweden, Denmark, Norway, Finland, the Netherlands, Italy and Spain (Pinnelli and De Rose 2001). Indeed delayed first child birth is now considered one of the most characteristic features of the most recent fertility change, a fact which has lead some authors to talk of yet another transition: the "postponement transition toward a late-childbearing regime" (Kohler, Billari and Ortega 2002).

Spain is a prime case here, since Spanish women are now the oldest first-time mothers in Europe, and arguably in the world, bearing their first child on-average at around 29. The Spainish case, however, is by no means unique. Women in at least six other European countries (Ireland, Italy, France, the Netherlands, Sweden, and Switzerland) and in Japan currently have their first child at an average age of over 28.

At the other extreme, in a number of OECD countries a significant minority of women still have their first child as a teenager and the persistence of an early-childbearing subculture contrasts with the more general phenomenon of ever-later childbearing among the majority of women. Such 'birth-polarisation' can now be found in a number of countries, but it has reached pronounced proportions in Ireland, the United Kingdom and, outside Europe, in the United States. It still remains unclear what exactly is driving the phenomenon, whether, for example, it is a result of the erosion of explicit norms relating to the timing of parenthood, of choice, of constraints, of careful planning, or of unanticipated ‘accidental’ pregnancies or whether it is the result of a parallel process of 'polarisation' of incomes and opportunity. (Sobotka, 2004)

The mean age at first birth in many 'late first birth' countries is, in fact, likely to continue to increase further in the near future, and may well eventually come close to the 30 year benchmark. However a variety of factors have lead some demographers to suggest that it unlikely that the mean age will increase notably above this threshold (at least in the short-term given current assisted reproduction technologies). Sobotka (2004) concludes that a mean age of 32 appears to constitute something of an outer limit for the record late-first-birth pattern. As he indicates, only if first births among young women under 25 were to be virtually eliminated could the mean age at first birth increase much further, and then only to the enter the 32-34 age interval.

There is one European society where the MAFB currently appears to have stabilised, at least temporarily, at a comparatively high level, and that is the Netherlands, where first birth postponment seems to have come to an end end around 1998. Curiously, as Sobotka notes, the Netherlands is also a country where demographers, other researchers, and even some politicians and journalists, have vocally expressed concerns about the consequences of late parenthood (e.g. van Nimwegen et al. 2003). However it is still too early to decide what importance to give to this fact - or how stable the pause in upward age movement actually is. However the Netherlands example should be followed closely as it may provide evidence regarding at least one line of possible policy initiative.

Another important feature of the ongoing process of birth postponement is its duration. The process has been operating now for significant periods of time (extending for over three decades in many cases) and such continuity stands in marked contrast with the other two major shifts in fertility tempo that occurred in industrialised countries in the 20th century: the fertility delay which occured during the economic depression of the 1930s, and the fertility advancement that took place during the 1950s and the 1960s. The latter is of course well known for the part it played in generating the major post-war baby booms.

Kohler, Billari, and Ortega (2002) point out that the current postponement shift has in and of itself many of the features of a ‘transition’, in this case one from a relatively-early to a relatively-late first birth timing pattern. This view seems to imply the idea of self-perpetuation, of a built-in momentum which continues to drive first birth delays over a long period of time irrespective of actual socio-economic conditions (Kohler, Billari, and Ortega 2002: 664) and this sets it apart from other earlier tempo shifts, which were, as we have seen, conjunctural or 'shock-like' in their incidence. The transition idea also implies the existence of an end state, but it is far from clear that any such end state exists, or that there is any ultimate lower bound to fertility.

Sobotka describes how,in Europe the recent postponement of first births has taken place in three main waves(Sobotka, 2004). During the first wave (1971-73), the shift towards late motherhood was initiated in eight countries of Western and Northern Europe. Similar changes occured in the United States and Japan at around the same time. The second wave (1980 - 85) included the Southern European countries (with the exception of Italy, where the process appears to have commenced around 1977), two peripheral 'outlier' countries in Western Europe (Austria and Ireland) and several Central European ones (East Germany, Croatia, Hungary, and Slovenia). The last wave (1992-1995) began shortly after the fall of the Berlin wall and included almost all the remaining post-communist societies. Only the former republics of the Soviet Union - Belarus, Moldova, and Ukraine - failed to commence their increase in MAFB until the final years of the 1990s.

Sobotka goes on to argue that there is an underlying structure to this fertility change. Initially there is a sudden and notable decline in births, and the appearance of this phenomenon may be considered an indication of the onset. Then there is a protracted period of very low TFR levels, and finally there is a ‘catching-up' stage as the postponment reaches completion and TFR levels begin once more to rise above the earlier 'lowest-low' level. The whole 'transition process' lasts at least two decades, since following the change in timing TFRs normally need a minimum of 15 years to attain a new equilibrium in conformity with the new set of age-parity birth probabilities. The magnitude of the eventual fertility ‘recovery' varies in a way which reflects both the underlying quantum change as well as the speed of the original birth decline (Sobotka, 2005).

For those who argue that the postponement of first births is one of the defining characteristics of a 'second demographic transition' (Lesthaeghe and Neels 2002: 333) - as opposed to constituting a distinct, ‘third transition’ (Kohler, Billari and Ortega 2002: 664) - its initiation may be thought to serve as a suitable indicator for the onset of this transition.

At the macro level, evidence concerning the relation between first birth postponement and eventual low fertility is still fairly controversial (Frejka and Calot 2001b, more references here) - indeed it could be said that the very idea of lowest-low fertility itself remains a controversial one given the fact that it owes, in part, its existence to the tempo effect.

So where exactly are we? Well, the evidence is indeed contradictory. In some countries - for example Northern Europe, Belgium - the first-birth delay seems to have involved a 'pure' postponement effect without apparently large consequences for eventual fertility levels, since the catching-up process is normally sufficient to cancel out a large part of the previous impact (although non of this happens without producing significant damage to the age pyramid in the process, as Lutz points out). Replacement level fertility is in no case, however, attained, and the more common outcome is a final achieved cohort TFR in the 1.7 - 1.9 range. In other countries – e.g. France and Switzerland - cohort fertility levels seem to be affected only slightly. Finally in Southern and Eastern Europe, Germany and Japan the evidence to date suggests that a consistent and sustained contraction of cohort fertility appears to follow the postponement of first birth for cohorts born after 1960.

Be this as it may, the empirical evidence indicating that something more than a ‘pure’ postponement process is at work is mounting and micro-level analyses have begun to offer confirmation that shifting motherhood to ever later ages is associated with a reduction of completed fertility (Morgan and Rindfuss, 1999; Billari and Kohler 2000). Kohler, Skyitte and Christiansen (2001), using longitudinal data from a sample of Danish twins, find, for example, that for every year by which the first childbirth is deferred, there is a reduction of 3% in the number of children achieved by the end of the reproductive life. Kohler, Billari and Ortega (2002) estimate that for Italy and Spain the postponement effect is even higher: in fact, each year of delay implies a reduction of completed fertility of between 2.9 and 5.1%. A weaker effect, declining across cohorts, has been observed in Northern Europe and in the US (Morgan and Rindfuss 1999).

What Drives Postponment?

A large number of tentative 'causes'have been offered in an attempt to explain postponement. Some of these stress economic and structural constraints (e.g. Happel et al. 1984), while others emphasise the importance of socialisation-process changes within the family, and the transition in social values (e.g. Schizzerotto and Lucchini 2002, Lestaeghe 1995).

Broadly speaking the economic approaches (see Livi Bacci and De Sanctis) assume that reproductive behaviour is the outcome of a rational choice process and that individuals have almost complete control over fertility. This leads many authors to hypothesise that the timing choice requires an evaluation of the costs and benefits associated with parenthood from a long run, as well as from a short run, perspective.

Postponing motherhood in circumstances of uncertainty - uncertainty surrounding, for example, future economic possibilities and future union stability - allows those who do this to dispose of an opportunity for a more precise evaluation of the costs and benefits of childbearing as well as of future earnings prospects. There is, if you like, a sort of options advantage attached to postponement (Kohler, Billari and Ortega 2002, Simò, Golsch and Steinhage 2002).

However, delaying childbirth is not a cost-free decision. As desired age at motherhood goes up women approach their natural biological age limit. The implications of delayed childbearing are complex. On the one hand, recent medical advances provide many women with the opportunity to 'sequence' (Blair-Loy, 1999; Garey, 1999), and achieve career success prior to the raising of children. For many career women in earlier times, this option was not available and they were often forced to remain childless in order to stay on a given career track. On the other hand, the risks associated with delayed childbearing are not, it seems, that widely known. Thus the approach of birth ages towards previous biological limits increases demands for medically assistanced reproduction - e.g. in vitro fertilization - and raises the anticipated level of biomedical cost (Wetzels, 1999, ch. 7). In addition, late mothers are subject to more substantial risks for their own health and for their late-born child (Gustafsson, 2001).

There are various known maternal risks associated with childbearing. As the mother’s age increases, a point is reached where there is a dramatic increase in the likelihood of childbearing complications. In the first place, the probability of carrying a child to term is significantly reduced after the age of 35. Cohen & Sauer (1998) found, for a sample of pregnant women receiving no infertility treatment, that spontaneous abortion rates rose from 10 percent in the 30-34 age group to almost 30 percent in the 35-39 one. According to Cohen and Sauer, even after fetal heart monitoring, miscarriage rates were still significantly higher in women of 35 and older. (More references from the postponement conference).

Gilbert, Nesbitt, & Danielson (1999) found that women age forty or over have a higher chance of requiring assisted childbirth: forceps, cesarean, and vacuum deliveries occur at a rate of 61% for this group, whereas younger women only have a 35 percent risk of needing such assistance. According to Scholz, Haas, & Petru (1999), in women over age 40, induction of labour was higher than for controls, while maternal morbidity due to childbirth is also found to increase with maternal age over 35 (van Katwijk & Peeters, 1998).

In addition, there are a number of fetal risks associated with advanced maternal age. In a study of 379 mature prospective mothers of 35 and older, who were compared to a control group of 379 prospective mothers, aged from 20 to 30, there were five stillbirths in the mature age group and none in the other younger age group (Barton, Bergauer, Jacques, Coleman, Stanziano, & Sibai, 1997). Furthermore, there is a significant difference in unexplained fetal death in pregnant women over the age of 35. According to Fretts & Usher (1997), one in 440 fetal deaths were unexplained in women 35 and older, whereas, one in 1000 fetal deaths were unexplained in women younger than 35.

Traditionally our reproductive biology has meant that female fertility has peaked in the mid 20s, followed by a steady then substantial fertility decline as the women moves through the 30s and 40s (Schmidt-Sarosi, 1998). Women over the age of 40 account for only one percent of all live natural births, a rate that falls to 0.01 percent for women over the age of 47 (Nesbitt, Bythell & Redfern 1999). These differences in childbearing are, in turn, linked to an increasing probability of infertility as a women ages. Griffin & Panak (1998) found that women under the age of 40 using in vitro fertilization and gamete intrafallopian transfer had a 20 percent success rate, whereas women over 40 had only an eight percent success rate. A study by Schmidt-Sarosi (1998) looking at all assisted reproductive technologies found that women younger than 35 had a 25.3 percent live birth rate, women 35-39 had a 18.2 percent live birth rate, and only an eight percent success rate was found in women 40 and older.

It is evident then that postponement decisions which result in unwanted childlessness constitute a good example of informational deficiencies, and the macro economic consequences (for eg lifelong savings), as well as the individual life-history consequences may well be important.

Paragraph above: life rescaling, does this affect our reproductive biology?

This next paragraph doesn't belong here.

In general higher levels of education are found to have a double effect on age of first birth: it leads to its postponement and lowers the probability that it may ultimately be realised (Nicoletti and Tanturri, 2005). However it should be noted that more highly educated women, once they do go down the road of motherhood, have a relatively high probability of having more than one child. Rendall and Smallwood (2003: 24), for example, found that in England and Wales 90.8% of higher educated women born in the 1954-1958 cohort and who gave birth to a first child at age 30 eventually gave birth to a second child. Nicoletti and Tanturri also find a very strong relationship between timing of first birth and age at the onset of the work career for all EU countries, finding that women, after first job commmencement, wait an average (depending on country) of between 3 and 7 years before deciding to have their first child.

Explanations For The Postponement/Factors Driving the Second Stage

There are in fact several competing theories of how fertility decisions are taken available in the literature. However a broad set of determinants are shared across most of the accounts of recent fertility declines in OECD countries. These determinants (or causal factors) include: (Elaborate a bit on the theories).

i) the perceived material and psychological benefits provided by children;

ii) the direct and opportunity costs of children which are incurred by their parents;

iii) the broad economic environment in which reproductive decisions are taken: the labour market insecurity and uncertainty faced by young people, changes in women’s economic roles, and the increased valuation of women’s work;

iv) a series of individual lifestyle factors, such as grater values attached to autonomy and self-realisation, greater willingness by women to adjust family aspirations to pursue career goals, and the diffusion of alternative forms of relationships;

v) changing societal and cultural norms, such as those which identify the division of home responsibilities within families and those underpinning the functioning of the welfare and tax systems.

Lesthaeghe (2001: 17-18) divides the relevant factors into two groups: one general and one country specific. In so doing he offers the following inventory of factors that help account for the new patterns of family formation and postponed parenthood:

General factors:

1) increased female education and female economic autonomy;

2) rising and high consumption aspirations that created the need for a second income in households and equally fostered female labour force participation;

3) increased investments in career developments by both sexes, in tandem with increased competition in the workplace;

4) rising ‘post-materialist’ traits such as self-actualisation, ethical autonomy,

freedom of choice and tolerance for the non-conventional;

5) a greater stress on the quality of life with a rising

taste for leisure;

6) a retreat from irreversible commitments and a desire for maintaining an ‘open future’;

7)rising probabilities of separation and divorce, and hence a more cautious ‘investment in identity’.

Country-specific factors:

1) the geographical mobility of young adults in tertiary education;

2) lack or availability of state subsidies for students in the forms of fellowships, housing facilities and transportation

subsidies;

3) the flexibility of the labour market, including the possibilities for part-time work;

4) youth unemployment;

5) minimum income guarantees;

6) costs and availability of housing, both for ‘starters’ and for

households in later stages of family formation (often linked to the structure of labour market and its regulations);

7) contraceptive availability and methods mix; access to abortion.

Principal Causal Factors

It is perhaps possible to identify a small number of principal causal factors.

1/. Contraception

Views relating to the importance of contraception in the fertility decline have been divided. Some - pointing to the fact, for example, that the European fertility transition was achieved using coitus interruptus - treat it as being of relatively secondary importance (Pritchett, 1994; Gertler and Molyneaux, 1994; Schultz, 1994). However in the context of the post-1970 trend in delayed childbearing in the developed countries (the so called second transition) there can be little doubt as to its importance. Approaches to the question seem to fall into two main groups: (i) contraception is seen as a technical factor addressing a pre-existing demand for birth control; (ii) contraception has its own dynamic and independent effect on fertility postponement.

The birth pill has been particularly important in this context for two main reasons. Firstly it has shifted control over pregnancy completely onto the woman herself, and secondly, in and of itself, it constitutes one of the most efficient contraceptive methods ever to become available. Perhaps as a result of this the pill "consistently heads the list of things which have most changed women’s lives" (de Guibert-Lantoine and Leridon 1999: 91).

In countries where legislative and societal attitude changes facilitated the rapid spread of the pill, its wide use was closely linked with the early onset of the fertility postponement. Santow and Bracher (2001: 359) associate increasing contraceptive prevalence, the drop in teenage pregnancies, and the subsequent onset of fertility postponement in Sweden with progressive changes in the sex education curriculum and the free provision of contraception to young people in the mid-1970s.

In Spain a dramatic increase in the pill use took place after 1978 when the ban on contraception was removed: “it is within this modern contraceptive regime that women started to delay first birth” (Castro Martín 1992: 232). As a result, rapid postponement of fertility - as captured by the increase in the mean age at first birth - started in Spain in 1980. In the Czech Republic, the very intensive postponement of first births after 1992 progressed hand in hand with the rapid diffusion of pill use (Svobotka, 2005). Elsewhere, increased contraceptive use has enabled young married women to delay entry into motherhood through prolonging the interval between marriage and the birth of the first child (Blossfeld and Huinink 1991 for West Germany, Castro Martín 1992 for Spain, and Murphy 1993 for England and Wales).

Further indirect support for the effect of modern contraception on fertility postponement comes from evidence on undesired and ‘mistimed’ pregnancies and first birth timing. From a cross-country perspective, there appears to be a link between inadequate contraceptive use, especially among teenagers, a high proportion of ‘unwanted’ or ‘mistimed’ pregnancies among (very) young women, and lower age at first birth. In Western European England and Wales stand out for their high teenage fertility rates and relatively earlier timing of first birth, but globally the clearest case is that of the United States.

In the United States a very high proportion of unintended births, which still accounted for around a third of the total fertility rate in the 1990s (Frejka 2004), went hand in hand with low contraceptive use and high fertility rates among teenagers (Morgan 1996 and the sociological perspective of Furstenberg 2002) and a relatively early age at first birth (by far the lowest among the ‘Western’ industrialised countries).

A similar combination of low contraceptive use, high teenage fertility, and earlier first birth timing is found in a number of Eastern European countries (Svobotka 2005, Chapter 7). One could argue that increased contraceptive availabilty leads to reduced rates of unintended pregnancies and births, especially among teenagers and young adults, and that this in itself facilitates fertility postponement.

The case of Italy, where very low levels of period fertility and rapid postponement of first births were achieved despite relatively low use of modern contraception and low abortion rates seems to provide a counter-argument to the importance of pill diffusion (Delgado Pérez and Livi-Bacci 1992; Castiglioni, Dalla Zuanna, and Loghi 2001). However it is assumed in the literature that a high prevalence of traditional birth control methods, particularly withdrawal, was already present in Italy and researchers generally conclude that by 1996 the use of modern contraceptive technology (the condom and in particular the pill) had almost completely replaced coitus interruptus among young people and sexually active women not living with a partner. In general, post 1996, the use of contraception became common from first sexual intercourse onwards (Dalla Zuanna, de Rose, and Racioppi 2001).

It is apparent that the diffusion of the pill has changed the nature of decision-making about childbearing. Surveys of sexual and reproductive behaviour indicate that women increasingly adopt pill-use at an early age, often from the start of their active sexual life (see e.g. de Guibert-Lantoine and Leridon 1999 for France; de Graaf and Lodewijckx 2000 for Flanders and the Netherlands; Dalla Zuanna, de Rose, and Racioppi 2001 for Italy; Bajos and Guillaume 2003 for a cross-country comparison). Continuous pill use has become the norm; in most countries women can also prevent unwanted pregnancies and births with the use of abortion and emergency contraception (Bajos and Guillaume 2003).

Here it is possible to identify another important break with the contraceptive patterns prevailing until the 1960s (or later in many countries), since initially contraceptive efforts were mainly focused on preventing additional pregnancies after a couple reached their desired family size. However this is certainly no longer the case and Beets et al. (2001: 21) have aptly drawn our attention to the way in which in the Netherlands modern contraception has changed the perception of sexual relationships within partnership from one of ‘getting children’ (kinderen krijgen) to a decision-making process over whether or not to ‘take children’ (kinderen nemen).

2/. Changing Household Structure

Van de Kaa (1994) links the spread of modern contraception with the change in living arrangements which characterise the postponement process: "the freedom from unwanted pregnancy has led to a new contextual or mental model of sexual relationships and of the connection between procreation and partner relations." Van de Kaa has asserted that ‘perfect contraception’ paradoxically played a key role in the spread of cohabitation and the subsequent postponement of marriages, in the decline of marriage rates, and in the increase in extra-marital childbearing.

The changing nature of intimate relationships has produced a profound transformation in family forms and living arrangements, especially since the late 1960s. The main contours of the changing European family are well known (e.g. Kuijsten 1996; Kiernan 2002) and have become the cornerstones of the concept of the second demographic transition (see e.g. Lesthaeghe and van de Kaa 1986; Lesthaeghe 1995; Lesthaeghe and Surkyn 2002; van de Kaa 1987, 1994, 1997, and 2001). Marriage has been increasingly replaced by cohabitation, extended spells of single living, and unconventional living arrangements.

Accelerating divorce rates have further eroded marital unions. The waning of marriage as an institution has led to the pluralisation of families and living arrangements. Marriage has also ceased to be the only socially accepted pathway to childbearing. Sex has been separated from reproduction, and reproduction has been detached from marriage. The separation of sex from procreation was made possible by the diffusion of the contraceptive pill (see the previous section).

3/. Changing Value Set

Societal norms, in particular with respect to gender roles, have a profound influence on fertility. If there is an "incoherence" between women’s roles within the family, the working place and society at large, fertility may well be negatively affected.

Societal norms as to the ideal, or standard, family size may also influence fertility. In societies where large families become less frequent or less socially accepted individuals possibly lower their preferred family size away from the previous family norm. In societies where out-of-wedlock births are less accepted fertility may wellalso be discouraged.

The 'traditional' explanation of women’s reproductive and childbearing decisions relies on the "rational choice" approach. This model of fertility, which was pioneered by Gary Becker (1981) in the 1970s and 1980s, regards individual decisions on having a child to be the result of a utility maximization process influenced by the economic cost and benefits of children, and subject to income constraints and individual’s preferences. Within this framework, the decline in fertility that characterises the developed countries may be the consequence of the higher price of children relative to other goods, of lower family incomes, or of a change in preferences for having children relative to other consumption goods. This model, which has been very influential in the literature, also lies at the core of most policies aimed to influence women’s childbearing decisions. For example, reductions in the cost of children (e.g. as a result of public subsidies) or increases in the income of women of reproductive age (e.g. due to higher transfer payments) would be considered to increase demand for children. Becker’s basic model has been extended over time in an attempt to account for additional aspects of the childbearing decision, such as 'quality' of children and timing of births (Hotz et al., 1999).

One rival to the Becker model is the options theory one. This is frequently associated with a tradition of life course analysis.Much of the life course analysis concentrates on the detection and documentation of a structure in the pathways of life (Willekens 1999: 26). Individual behaviour is conditioned by various social constraints, determining the options available for each individual actor; thus, any behaviour at the micro level is influenced by a particular context- a feature labelled within the theory as embeddedness (ibid.: 28).

On most accounts of the postponment process individuals are seen as rational actors trying to utilise information available to them to achieve their goals within various constraints, such as material resources, institutional regulations, and cultural norms.

Decision-making on important life course transitions is strongly influenced by an interplay between the various domains of life, and this leads to potential conflicts between the various available parallel careers. In general the decision to have a child is thought to depend strongly on current and expected partnership status, employment, and educational pathway decisions (e.g. Willekens 1991b).

In modern societies rather than strict ‘sequencing norms’, what exist are a number of commonly shared concepts of what constitute the necessary preconditions for parenthood: leaving the parental home, finishing education, and accumulating resources. Leaving the parental home in fact constitutes in most societies a precondition for making individual choices on union formation and parenthood (Billari, Philipov, and Baizán 2001: 354).

Later union formation, less stable employment and living arrangements, and higher expectations of partnership quality mean that many modern young men and women hesitate in committing themselves to parenthood. An increasing feeling of entitlement to experiencing a variety leisure activities and acquiring a variety of consumer products prior to entering parenthood also presents an additional obstacle fuelling fertility postponement (e.g. Presser 2001).

At the same time the social definition of ‘being too old’ to have children has been modified in recent decades in favour of a higher acceptability of late childbearing (Rindfuss, Morgan, and Swicegood 1988). The evidence suggests that there are generally shared norms on a preferred ‘timetable’ for having a first child, and this timetable tends to move towards ever later ages paralleling the trend of delayed parenthood, during this process the 'childbearing years' come to encompass a relatively narrow age interval. (Toulemon and Leridon 1999, France.... Matsuo 2003: Japan).

Some authors have also identified the existence of comprehensive pension systems as an important determinant of low fertility in developed countries (Livi Bacci, 2001b). While, in traditional societies, a higher fertility rate may reflect the parents wish to have support available from children when they reach old age, the expansion of alternative forms of support for the elderly might have reduced the importance of this factor.

This hypothesis would suggest that countries with more generous pension systems, and with expectation that this generosity will continue in the future, will also experience lower fertility. As the income of older people does not depend exclusively on the social security system, Sleebos used the mean disposal income of those over 65 (relative to income available to the working age population as a whole) in an attempt to identify a more general indicator of the economic resources of the elderly in order to examine any possible relation. He found there was a negative relationship between higher relative income in old age and total fertility, although the relationship was not a strong one (Sleebos, 2003).

Those who employ the concept of a second demographic transition tend to see late childbearing as the result of a fundamental social, economic, and cultural transformation, one which is defined as much by changed norms related to the family and parenthood as it is by decision-making on the timing of childbearing. On this view the purely economic perspective on family and childbearing, as epitomised by Becker (1991), is thought to provide too narrow a view insofar as it ignores the altered character of norms and attitudes in relation to the family and reproduction.

4/. Costs of children.

The costs of having children may be considered to be consituted by all the child-related expenses faced by families with children. While difficult to quantify, these costs may have increased in recent decades relative to the prices of other goods and in particular technological products. Since parents feel an increasing desire and obligation to invest more heavily in the education of their children these costs rise even further. Costs related to providing appropriate housing for families with children may also increase as urbanisation unfolds, since in large cities housing costs tend to increase with space, and cheaper forms of housing may be considered to be less suited for raising children. Another reason for the higher costs of having children is the tendency for couples to increasingly live at an ever greater distance from their parents, with the resulting need to rely on costly child-care facilities. Beyond direct costs, mothers also incur opportunity costs due to earnings-loss during absence from work, and to the impact of work interruption future career development. The importance of these opportunity costs will partly depend on the feasibility of combining paid work with child-rearing (with opportunity costs being higher the more the difficulty that is associated with the work-mother combination).

5/. Employment and Education

During the last four decades young Europeans have spent an ever larger portion of their lives in education. In particular there has been a massive expansion in post-secondary education. Post-industrial economies generate a growing demand for an ever more highly educated and flexible workforce. For individuals, pursuing higher education constitutes the main pathway on-route towards finding a stable job, receiving sufficient wage, and increasing career prospects (Kohler, Billari, and Ortega 2002). Most men and women continue to be enrolled in some form of education in early adulthood, and the expected number of additional years in post compulsory education extends as far as 8.5 years (in France, OECD 2001).

Consequently, up to half of all those in the 20-24 age group (and even up to between 50 and 55% in the cases of Denmark, Finland, and France) are enrolled in full-time or part-time studies. Women in particular have availed themselves of this opportunity, and they now form more than half of the total number of graduate and postgraduate students in a majority of European countries.

Such educational expansion has direct implications for fertility trends. In present-day Western societies, years spent in education are almost universally perceived as incompatible with family formation. This view has been supported by a number of studies which have illustrated that ‘being in education’ strongly reduces the probability of having a first child (see e.g. Rindfuss, Morgan and Swicegood 1988; Blossfeld and Huinink 1991; Kravdal 1994; Blossfeld 1995; B. Hoem 2000; Baizán, Aassve, and Billari 2003).

Beets et al. (2001) found that highly educated women were the ‘forerunners’ of the childbearing postponment in European countries, and it has been estimated that an increase in education level can explain about half the increase in the mean age at first birth found between Dutch women in the 1931-40 and the 1961-65 cohorts. Substantial differences in first birth timing according to education levels are found across all developed societies. The example of French women born in 1950-55 is illustrative: those who left education after completing primary school gave birth to a first child at an average age of 22.6, while those with four and more years of post-secondary education had their first child at an average age 28.4 (Meron and Widmer 2002: 303).

Higher education both enhances the labour market status of the individual labour and increases the ‘opportunity cost’ of childbearing. People with higher education have values and preferences which are distinct from those with less education, although this may not necessarily be due to the effects of the education itself, as (self-)selection to higher education may reflect value orientation as well as contibuting to forming it. Education goes hand in hand with increased material and career aspirations. Individuals with higher education are more proficient in obtaining and assimilating information, and are less sensitive to social pressure (Bouwens, Beets, and Schippers 1996). They embrace values such as autonomy, independence, and self-realisation.

Educational groups may be viewed as "possessing a cultural capital, a 'Weltanschauung', and a preference map of some stability at ages around which the transition to adult life (…) is centered" (Lesthaeghe and Surkyn 1988: 17). Thus, the reluctance to start a family among women and men with higher education could be seen as reflecting increased resistance to normative pressures, higher levels of flexibility, and stronger attachment to career building, as well as an anticipation of higher costs being associated with having children and an indication of a higher valuation of individual autonomy (e.g. Liefbroer 1998).

Traditionally employment and motherhood were seen as incompatible roles. This view is supported by the economic perspective on family formation, especially by Becker’s (1991) argument that the increasing earning power of women increases both their labour force participation and the opportunity costs of childbearing. This perspective is subsequently developed to incorporate the idea of education implying a lower demand for the ‘quantity’ of children, since educated women prefer to invest more in the education and other training of their children (Becker 1991: 153).

Many recent studies, however, have found that the relationship between labour force participation and fertility is not a straightforward one, and may be filtered through a number of additional factors. Liefbroer and Corijn (1999) advocate a dynamic perspective on the compatibility between family life and labour-force participation, stressing that societal differences, cohort membership, age, and education level modify this relationship.

From a macro perspective, a seemingly counter-intuitive positive association between period total fertility and women’s labour participation has been found in OECD countries since the mid-1980s (see Brewster and Rindfuss 2000; Rindfuss, Guzzo, and Morgan 2003; Billari and Kohler 2004; Engelhardt and Prskawetz 2004). The complex relationship between the two interdependent ‘careers’ of employment and fertility (see Willekens 1991b) considerably influences the timing of family formation. Brewster and Rindfuss (2000: 282) noted that birth timing and spacing "may comprise key components of strategies to balance work and family responsibilities."

The difficulty that women face in combining work and family responsibilities account for the fact that, in general, the proportion of women with children is higher among those who do not work than it is among those who do. Large differences however are to be found across countries, both in terms of the gap between these two groups of women (the gap being much larger in Canada, the Netherlands and New Zealand) and in the proportion of women with children by employment status. The type of job and contract held by the mother is also important when it comes to decisions about having children. Part-time jobs, in particular, generally allow women greater opportunities to combine work and family responsibilities. Across countries, the proportion of women with children is generally higher among those working part-time than among those working full-time, with differences between the two groups being larger in the Netherlands, New Zealand and Sweden but much smaller in France, Italy and Spain (Sleebos 2003).

Because part-time employment is often the preferred option of the working mother, in those countries where part-time employment is rare women will have to choose between either leaving work and having children or taking up a full-time job, neither of which is likely to be their preferred option.

Additional information on how women’s reproductive and employment decisions interact is provided by labour force survey data (OECD, 2001). Trends in women’s employment rates suggest that while these have increased in almost all countries (OECD, 2002) those for women with young children remained, with few exceptions (Ireland, Luxembourg, the Netherlands, the United States) fairly stable throughout the 1990s (OECD, 2002).

6/ Uncertainty

Mills and Blossfeld, for instance, (Mills and Blossfeld, forthcoming: 18-19) distinguish between various types of uncertainty.

(1) economic uncertainty, related to the "economic precariousness of an individual’s employment and educational enrolment circumstances"

(2) temporal uncertainty, and

(3) employment relationship uncertainty, reflecting the type and precariousness of one’s employment contract.

In the Mills and Blossfeld framework, being unemployed leads to a high level of individual economic uncertainty, whereas rising unemployment rates could lead to a more generally higher temporal uncertainty. Young adults are increasingly susceptible to a variety of forms of uncertainty, especially with regards to their employment situation, and this has a disproportionate impact on disadvantaged - and especially on less-educated - social groups. In addition, the accelerating pace of change, the increased unpredictability of social and economic developments, and the increase in the flow of available information all serve to create heightened uncertainty attached to both possible behavioural outcomes and to the relative probability of these outcomes. To this general uncertainty may be added doubt about the amount of information which needs to be collected for taking any particular decision (Mills and Blossfeld forthcoming: 17). The authors attribute rising uncertainty among young adults to the broadly defined forces of globalisation.

7/. The Family

One factor stands out in the present context above all others: the presence of a prevailing conservative (Continental) welfare regime or a ‘familiastic’ (Southern European) welfare model appears to have a negative impact both on the tempo (i.e. it tends to induce additional postponement) and on the quantum of fertility (by reducing the opportunities for women to have an independent career and have multiple children, e.g. Esping-Andersen 1999). High unemployment rates coupled with low family benefits, policies supportive of the traditional male breadwinner family model, costly and scarce childcare, expensive rental housing, and limited part-time work opportunities all serve to characterise many societies with very low fertility and very late timing of motherhood.

Moreover, norms prescribing stay-at-home mothers, and in particular those which stress the desireability of this in the presence of young children, tend to increase role incompatibility between employment and fertility (Brewster and Rindfuss 2000). Southern European countries, especially Italy and Spain, exemplify just such a situation (see Baizán, Michielin, and Billari 2002 for Spain and Dalla Zuanna 2001 for Italy) and the extremely low levels of fertility to be found in these societies are often attributed to the strength of this role incompatability.

The McDonald Thesis

In recent decades, in modern industrial economies, participation of women in the formal labor force has expanded rapidly. This tendency, reflecting market forces but also encouraged by government policy (partly as an antidote for deteriorating dependency ratios as the population becomes older), is likely to continue. Among the many factors advanced to explain the generally low level of fertility despite general material affluence, observers have pointed to the double burden on women of both raising children and working outside the home. To the extent that higher birth rates are seen to be socially desirable, the derived policy prescription would seem to be the adoption of measures that make motherhood and women’s labor force participation more compatible.

Comparatively higher fertility levels in those countries (most notably in Scandinavia) where measures to increase this compatability have been systematically applied, when compared to countries (especially those in Southern Europe) where they are largely absent, suggest that enhanced compatibility (through day-care services, flexible work-hours, liberal sick-leave allowances, and the like) is an effective pronatalist policy even if motivated by other considerations. But it is far from clear whether the fertility differential so generated is high enough to bring the total fertility rate back to replacement level.

Peter McDonald, in a series of papers (2000, 2000a, and 2002), has suggested that very low fertility is an outcome of high levels of gender equity in an environment dominated by individual-oriented institutions (implying a relatively high structural compatibility), combined with a persisting gender inequality within the family. This situation poses a dilemma for women, who may increasingly perceive their family role as inconsistent with their individual aspirations.

Evidence for such continuing inequality within the family unit is to be found in numerous time-budget surveys, which clearly indicate that women still do most of the domestic work (e.g. Joshi 1998, Table 2; Esping-Andersen 1999: 57-60). Both among housewives and employed women, the time spent on ‘social reproduction’- child-rearing and related domestic tasks - is not compensated-for in terms of family benefits or other means, and an equal division of domestic tasks between men and women seems still to be a long distant goal. As Joshi (1998: 177) notes, in "the private arena of home life, at least among young adults, the ideology of sex equality runs ahead of the practice."

Within this context, McDonald has emphasised the importance of the 'coherence' between the levels of gender equity in different social institutions, and between the roles, functions and preferences of different actors. McDonald distinguishes between two types of gender roles, in paid and unpaid work. The first setting ('male breadwinner model') is characterised by a complementary division of labour between men and women, with a clear distinction between the responsibilities of men for jobs and market-income, and those of women for care and domestic work. The second setting (the 'gender equity model') is characterised by a symmetric division of responsibilities within the family and the labour market. While both settings may a-priori lead to a high fertility rate, societies where changes in women’s economic roles and aspirations are not matched by similar changes in institutions and family responsibilities seem destined to experience lower fertility rates.

Curiously, and contrary to what might be expected from McDonald’s original hypothesis, lower structural and cultural incompatibility do appear to go hand in hand. ‘Family-oriented’ welfare regimes in Italy and Spain, hindering a compatibility of employment and childbearing for women and embracing the traditional breadwinner model, are good examples of both (e.g. Dalla Zuanna 2001, Del Boca 2002). In Esping-Andersen’s words (1999: 67) "the great paradox of our times is that familialistic policy appears counter-productive to family formation."

West Germany, a region which currently has the highest childlessness rates in Europe, also has a long-lasting shortage of day-care facilities and an institutional system which is conducive to reproductive polarisation, encouraging women with small children to stay at home and serving as an obstacle for those who wish to combine work and childrearing.

The relatively high level of childlessness in England and Wales has also been related to the incompatibility of motherhood and upper-level employment, resulting in a large proportion of higher educated women remaining childless (Ekert-Jaffé et al. 2002; Rendall and Smallwood 2003).

The rapid increase of childlessness in Italy and Spain also suggests that prevailing cultural norms may constitute an additional obstacle to parenthood. Italy is an interesting case in this respect, since it is a country where the various available theoretical frameworks offer quite conflicting signals as to the expected extent of future childlessness.

On the one hand, Italian society is characterised by a deep-rooted familistic culture (Reher 1998), which attaches a high value to in-family parenthood. However, in addition to negative socio-economic factors, such as a high unemployment rate among young people, the importance attached to the traditional family may also serve to explain the very late pattern of home leaving and partnership formation, as well as the low compatibility of child-rearing with female employment (Dalla Zuanna 2001, Micheli 2004).

Living in a welfare system that hinders individual autonomy, while being increasingly reluctant to marry and enter parenthood, young Italians commonly stay in the parental home into their early thirties (Billari and Rosina 2004). The lack of gender equity within the family in combination with the expanding economic opportunities for women outside the family sphere has been rightly identified by McDonald (2002) as a further obstacle to family formation, and clearly forms part of the explanation for very low fertility in Southern Europe. Livi Bacci (2001a), unsurprisingly, sees signs of a 'delay syndrome' (one which will unwind) in all of this, particularly in the context of the Southern European countries.

In the eastern and central European case, the collapse of the earlier communist regimes paved the way for increased labour market competition and unemployment, a scaling down of the previous social policies, an increase in the duration of education, a rapid spread of modern contraception and, in general, lifestyles that are less compatible with parenthood. Although the plummeting levels of period fertility in these countries have been frequently attributed to the broadly defined effects of economic and social uncertainty (e.g. UN 2000), survey data point to the pivotal role of the changing character of partnerships.

The Population Policies Acceptance Survey (PPAS), carried out between 2000 and 2003, indicates that childless men and women who do not intend to become parents, or who remain uncertain, most typically quote the lack of a steady partner as a reason. The costs of children and concerns about maintaining one’s living standards are quoted frequently as well, whereas relatively few men and women perceive their professional activities as a main reason for the intention to remain childless.

Besides being tightly linked to the current partnership status, childbearing intentions are also strongly connected to preferred living arrangements, revealing different lifestyle preferences. Respondents that favour other living arrangements than marriage, including those who prefer to cohabit first and marry later, are considerably less certain about their childbearing intentions. This finding is particularly helpful for explaining the projected sharp increase in childlessness in Poland, a fairly traditional and dominantly Catholic society. Although the younger generations have actively embraced the ‘reflexive’ model of partnership, and now tend to wait longer before entering a stable union, their search for the most suitable partner is hindered by the low societal acceptance of informal living arrangements. The PPAS results provide then a picture which is full of paradoxes.

Most recently, however, the predominant explanations of women’s reproductive decisions to be found in studies of chilbearing and fertility go beyond the focus on individuals’ decisions that characterised the earlier 'rational choice' models. Such studies typically stress the importance of cultural and institutional constraints within which individuals’ reproductive decisions take place.

On some accounts, emphasis is given to the 'risk aversion' of individuals when it comes to having children, and to the fact that both future costs and benefits of children cannot be known with certainty: when uncertainties about future economic, social or personal conditions increase, individuals may lean on the side of safety in order to avoid risk. In other accounts, emphasis is put on the emergence of post-material values in industrialised societies (such as individual self-realisation, satisfaction of personal preferences, and freedom from traditional forces of authority) and on changes in gender roles.

References

Bagavos C. and C. Martin (2001), “Low fertility, families and public policies”, Materialiensammlung, Heft 10, Austrian Institute for Family Studies, Wien.

Bajos, N. and A. Guillaume. 2003. “Contraceptive practices and use of abortion among adolescents and young adults in Europe”. In.: N. Bajos, A. Guillaume, and O. Kontula (eds.) Reproductive health behaviour of young Europeans. Vol. 1. Population Studies No. 42, Strasbourg: Council of Europe.

Becker, G. 1991. A treatise on the family. Enlarged edition. Harvard University Press, Cambridge.

Beets, G., E. Dourleijn, A. Liefbroer, and K. Henkens. 2001. De timing van het eerste kind in Nederland en Europa. Raport No. 59, NIDI, Den Haag.

Billari, F. C., D. Philipov, and P. Baizán. 2001. “Leaving home in Europe: The experience of cohorts born around 1960”. International Journal of Population Geography 7: 339-356.

Blair-Loy, M. (1999). Career patterns of executive women in finance: An optimal matching analysis. American Journal of Sociology, 104, 1346-1397.

Bloom, D.E., and J. D. Sachs. (1998)., ”Geography, Demography and Economic Growth in Africa”, Brooking Papers on Economic Activity 1998.

Blossfeld, H.-P. and J. Huinink. 1991. “Human capital investments or norms of role transition? How women’s schooling and career affect the process of family formation”. American Journal of Sociology 97 (1): 143-168.

Blossfeld, H.-P. 2005. "Globalization, Uncertainty and Fertility Decisions of Young People in Modern Societies", Paper presented at the Conference "Postponement of Childbearing in Europe", Vienna, Decemeber 2005.

Bongaarts, John. 2001. “Fertility and Reproductive Preferences in Post-Transitional Societies,” in Global Fertility Transition, a supplement to Volume 27 of Population and Development Review. Rodolfo Bulatao and John Casterline eds. New York: Population Council, pp. 260– 81.

Bongaarts, J. and R. A. Bulatao (eds.). 2000. Beyond Six Billion: Forecasting the World’s Population. Washington, DC: National Academy Press.

Bongaarts, J. and G. Feeney. 1998. “On the quantum and tempo of fertility,” Population and Development Review 24(2): 271–291.

Brückner, H., & Mayer, K. U. (in press). The de-standardization of the life course: What it might mean? And if it means anything. Wether it actually took place? Advances of Life Course Research.

Bumpass, L. L. and E. K. Mburugu. 1977. “Age at marriage and completed family size,” Social Biology 24(1): 31–37.

Caldwell, John C. 1982. Theory of Fertility Decline. New York: Academic Press.

Castiglioni, M., G. Dalla Zuanna, and M. Loghi. 2001. “Planned and unplanned births and conceptions in Italy, 1970-1995”. European Journal of Population 17: 207-233.

Castles F.G. (2003), “The world turned upside down: below replacement fertility, changing preferences and family-friendly public policy in 21 OECD countries”, Journal of European Social Policy, vol. 13(3), London

Castro Martín, Teresa, 1992, Delayed childbearing in contemporary Spain: trends and differentials, European Journal of Population/Revue européenne de Démographie (Historical Archive), Volume 8, Number 3, September 1992, Pages: 217 - 246

de Coninck-Smith, N, B. Sandin and E. Schrumpf (eds.), 1997, Industrious Children. Work and Childhood in the Nordic Countries 1850-1990, Odense.

Cunningham, H and P. P. Viazzo (eds.), Child Labour in Historical Perspective 1800-1985: Case Studies From Europe, Japan and Colombia, Unicef, 1996;

Dalla Zuanna, G. 2001. “The banquet of Aeolus: A familistic interpretation of Italy’s lowest-low fertility”. Demographic Research 4, Article 5: 133-161.

Dalla Zuanna, G., A. de Rose, and F. Racioppi. 2001. “Low fertility and scarce diffusion of modern contraception in Italy: a paradox to interpret”. Paper presented at the IUSSP Seminar “International perspectives on low fertility: trends, theories and policies”, Tokyo, 21-23 March 2001.

De Graaf, A. and E. Lodewijckx. 2000. “Fertility regulation”. In.: J. de Beer and F. Deven (eds.) Diversity in family formation. The 2nd demographic transition in Belgium and the Netherlands. European Studies of Population, Vol. 8, Kluwer Academic Publishers, Dordrecht, pp. 11-32.

De Guibert-Lantoine, C. and H. Leridon. 1999. “Contraception in France. An assessment after 30 years of liberalization”. Population: An English Selection 11, 1999: 89-114.

Delgado Pérez, M. and M. Livi-Bacci. 1992. “Fertility in Italy and Spain: The lowest in the world”. Family Planning Perspectives 24 (4): 162-171.

Demeny, Paul 1997. “Replacement-level fertility: The implausible endpoint of the demographic transition,” in Gavin W. Jones, Robert M. Douglas, John C. Caldwell, and Rennie M. D’Souza (eds.), The Continuing Demographic Transition. Oxford: Clarendon Press.

Dialog Project, 2006, More Details

Easterlin, R.A. 1987. Birth and Fortune. The Impact of Numbers on Personal Welfare. New York: Basic Books, 2nd edition.

Ekert-Jaffé, O., H. Joshi, K. Lynch, R. Mougin, and M. Rendall. 2002. “Fertility, timing of births and socio-economic status in France and Britain”. Population-E 57 (3): 475-508.

European Commission. 2001. Family and the desire for children, Eurobarometer 32, organised and supervised by DG Press and Communication, Brussels

Frejka, T. 2004. “The ‘curiously high’ fertility of the USA”. Population Studies 58 (1): 88-92.

Frejka, T. and G. Calot. 2001a. “Cohort reproductive patterns in low-fertility countries,” Population and Development Review 27(1): 103–132.

Frejka, T. and G. Calot. 2001b. “L’´evolution du calendrier des naissances par generation dans les pays `a basse fécondité `a la fin du XXe siècle,” Population 56(3): 397–420

Furstenberg, F. F., Jr. 2003. “Teenage childbearing as a public issue and private concern”. Annual Review of Sociology 29: 23-39.

Garey, A.I. (1999). Weaving work and motherhood. Philadelphia, PA: Temple University Press.

Gertler, Paul J. and John W. Molyneaux. 1994. “How Economic Development and Family Planning Programs Combined to Reduce Indonesian Fertility.” Demography. February, 31:1, pp. 33– 64.

Goldstein, J., W. Lutz, and S. Scherbov. 2003. Long-term population decline in Europe: The relative importance of tempo effects and generational length. Population and Development Review 29(4): 699-707.

Goldstein, J., W. Lutz, and M. Rita Testa. 2003. “The emergence of sub-replacement family size ideals in Europe”. Population Research and Policy Review 22 (5-6): 479-496.

Gustafsson S.S. (2001), Optimal Age at Motherhood. Theoretical and Empirical Considerations on Postponement of Maternity in Europe, Journal of Population Economics, 14, 225-247.

Happel S.K., Hill K.J., Low S.A.(1984), An economic analysis of the timing of childbirth, Population Studies, 299-311.

Hotz V.J., J.A. Klerman and R.J. Willis (1999), “The economics of fertility in developed countries. A survey”, reprinted from Handbook of population and family economics, Labor & Population, reprint series 99-08.

Jeon, Y. and M. P. Shields. 2005. The Easterlin hypothesis in the recent experience of higher-income OECD countries: A panel data approach. Journal of Population Economics 18(1): 1-13.

Kaplan, HS, Lancaster, JB, & Robson, A. 2003. Embodied Capital and the Evolutionary Economics Of the Human Lifespan. In: Lifespan: Evolutionary, Ecology and Demographic Perspectives, J.R. Carey & S. Tuljapakur (eds.) Population and Development Review 29, Supplement 2003, Pp. 152-182.

Kohler, Hans-Peter. 2001. Fertility and social interaction: an economic perspective, New York, Oxford University Press.

Kohler, H.-P., F. C. Billari, and J. A. Ortega. 2002. “The emergence of lowest-low fertility in Europe during the 1990s”. Population and Development Review 28 (4): 641-680.

Kohler, H.-P. and D. Philipov. 2001. “Variance effects in the Bongaarts-Feeney formula,” Demography 38(1): 1–16.

Kohler, H.-P. and J. A. Ortega. 2002. “Tempo-adjusted period parity progression measures, fertility postponement and completed cohort fertility,” Demographic Research 6(6): 91–144.

Kohler H.P, Skyitte A., and Christiansen K. (2001), The age of first birth and completed fertility reconsidered: Finding from a sample of identical twins, Max Planck Institute for Demographic Research, Working paper 2201-006, Rostock, Germany.

Lee, Ronald, 2003, Demographic Change, Welfare, and Intergenerational Transfers: A Global Overview, University of California, Berkeley, CEDA Papers.

Lesthaeghe, R. 2001. “Postponement and recuperation: Recent fertility trends and forecasts in six Western European Countries,” paper presented at the IUSSP Seminar on International Perspectives on Low Fertility: Trends, Theories and Policies, Tokyo, Japan, March 21–23.

Lestaeghe R. (1995), "The second demographic transition in Western countries: an interpretation" in Mason K.O. and Jensen A.M. (eds), Gender and family changes in industrialized countries, Oxford, Clarendon Press.

Lesthaeghe R. and G. Moors (2000), “Recent trends in fertility and household formation in the industrialized world”, Review of Population and Social Policy, No. 9.

Lesthaeghe, R. and K. Neels. 2002. “From the first to the second demographic transition: An interpretation of the spatial continuity of demographic innovation in France, Belgium and Switzerland”. European Journal of Population 18 (4): 325-360.

Lesthaeghe, Ron, and Johan Surkyn. 1988. Cultural dynamics and economic theories of fertility change, Population and Development Review, Vol. 14, No. 1, pp.1-45.

Lesthaeghe, R. and P. Willems. 1999. “Is low fertility a temporary phenomenon in the European Union,” Population and Development Review 25(2): 211–228.

Lindh, T. and B. Malmberg (1999). "Age structure effects and growth in the OECD, 1950-1990." Journal of Population Economics 12: 431-449.

LIVI-BACCI M., 2001a, “Demographic Shocks: The View from History”, in: Seismic Shifts: The Economic Impact of Demographic Change. Proceedings from the Federal Reserve Bank of Boston Conference Series No. 46

Livi-Bacci, M. 2001b. “Too few children and too much family,” Daedalus 130(3): 139–156.

Livi Bacci M. 2001c, “Comment: desired family size and the future course of fertility”, in R. Bulatao and J. Casterline (eds), Global Fertility Transition, Supplement of Population and Development Review, volume 27.

Lochhead C. (2000), “The Trend toward Delayed First Childbirth: Health and Social Implications”, ISUMA. Canadian Journal of Policy Research, volume 1, No. 2, Autumn.

Lutz, W., B.C. O’Neill, and S. Scherbov. 2003. Europe’s population at a turning point. Science 299: 1991-1992.

Lutz,W. Qiang Ren, Sergei Scherbov, and Xiaoying Zheng, 2005, "China’s uncertain demographic present and future", paper presented at the International Union for the Scientific Study of Population XXV International Population Conference, Tours, France, July 18-23, 2005

Lutz, W., W. Sanderson, and S. Scherbov. 2001. “The end of world population growth,” Nature 412: 543–545.

Lutz, W. and V. Skirbekk. 2004. How would ‘tempo policies’ work? Exploring the effect of school reforms on period fertility in Europe. European Demographic Research Papers 2004, No. 2. Vienna, Austria: Vienna Institute of Demography of the Austrian Academy

of Sciences.

Lochhead C. (2000), “The Trend toward Delayed First Childbirth: Health and Social Implications”, ISUMA. Canadian Journal of Policy Research, volume 1, No. 2, Autumn.

Lee, Ronald, 2003, Demographic Change, Welfare, and Intergenerational Transfers: A Global Overview, University of California, Berkeley, CEDA Papers.

Lutz, Wolfgang, Maria Rita Testa, Vegard Skirbekk, 2006. The "Low Fertility Trap" Hypothesis, Paper presented at the Population Association of America (PAA) 2006 Annual Meeting, March 30 - April 1, Los Angeles, California

Lutz, Wolfgang, Maria Rita Testa, Vegard Skirbekk, 2005. The "Low Fertility Trap" Hypothesis power point presentation at the Postponement of Childbearing in Europe conference held at the Vienna Institute of Demography, 1-3 December 2005, Vienna, Austria

Lutz, W., B.C. O’Neill, and S. Scherbov. 2003. Europe’s population at a turning point. Science 299: 1991-1992.

Lutz,W. Qiang Ren, Sergei Scherbov, and Xiaoying Zheng, 2005, "China’s uncertain demographic present and future", paper presented at the International Union for the Scientific Study of Population XXV International Population Conference, Tours, France, July 18-23, 2005

Lutz, W. and V. Skirbekk. 2004. How would ‘tempo policies’ work? Exploring the effect of school reforms on period fertility in Europe. European Demographic Research Papers 2004, No. 2. Vienna, Austria: Vienna Institute of Demography of the Austrian Academy of Sciences.

Macunovich, D.J. 2002, Birth Quake: The Baby Boom and Its Aftershocks. Chicago: University of Chicago Press

Macunovich, D.J. 2000, Relative Cohort Size: Source of a Unifying Theory of Global Fertility Transition? Population and Development Review, Volume 26 Issue 2, June 2000

Macunovich, D.J. 1998a, Relative Cohort Size and Inequality in the U.S. American Economic Review (Papers and Proceedings) May 1998 88(2):259-264

Macunovich, D. J. (1998) “Fertility and the Easterlin hypothesis: An assessment of the literature.” Journal of Population Economics 11:53-111.

Malmberg, Bo & Lena Sommestad. (2000). Four Phases in the Demographic Transition, Implications for Economic and Social Development in Sweden, 1820-2000. Arbetsrapport/Institutet för Framtidsstudier;Working Paper 2000:6

Marini, M. M. and P. J. Hodsdon. 1981. “Effects of the timing of marriage and first birth on the spacing of subsequent births.,” Demography 18(4): 529–548.

Matsuo, H. 2003. The transition to motherhood in Japan. A comparison with the Netherlands. Doctoral thesis, University of Groningen. Population Studies, Rozenberg Publishers, Amsterdam.

Mayer, K. U. (2004). Whose Lives? How History, Societies and Institutions Define and Shape Life Courses. Research in Human Development 1 (3), 161-187.

Mayer, K. U. 2001. “The paradox of global social change and national path dependencies: Life course patterns in advanced societies”. In.: A. E. Woodward and M. Kohli (eds.) Inclusions and exclusions in European societies. Routledge, London, pp. 89-110.

McDonald, P. 2005. "Fertility and the State: the efficacy of policy", paper presented at the XXV International Population Conference of the International Union for the Scientific Study of Population, Tours, France, July 18-23, 2005.

McDonald, P. 2002. “Sustaining fertility through public policy: The range of options”. Population-E 57 (3): 417-446.

Morgan, P. S. 1996. “Characteristic features of modern American fertility”. In.: J. B. Casterline, R. D. Lee, and K. A. Foote (eds.) Fertility in the United States. New patterns, new theories. Supplement to Population and Development Review 22, New York, Population Council, pp. 19-63.

Morgan, P. S. and R. R. Rindfuss. 1999. “Reexamining the link of early childbearing to marriage and to subsequent fertility,” Demography 36(1): 59–75.

Murphy, M. 1993. “The contraceptive pill and women’s employment as factors of fertility change in Britain 1963-1980: A challenge to conventional view”. Population Studies 47 (2): 221-243.

National Research Council. 2001. Diffusion Processes and fertility transition: selected perspectives, Committee on Population. John B. Casterline, (ed.). Division of Behavioural and Social Sciences and Education. Washington, DC, National Academy Press.

Nicoletti, Cheti, and Maria Letizia Tanturri, 2005, Differences In Delaying Motherhood Across European Countries: Empirical Evidence From The ECHP (European Community Household Panel Survey), Institute For Social And Economic Research, ISER Working Papers 2005-4

OECD (2001), Society at a Glance, Paris.

OECD (2002), OECD Employment Outlook, Paris.

Oeppen, J., and J. W. Vaupel. 2003. Broken limits to life expectancy. Science 296:1029-1031.

Pinnelli A., De Rose A. (2001), Delayed Fertility in Europe: determinants and consequences, paper presented at EAPS Population Conference 2001, Helsinky, Finland.

Presser, H. B. 2001. “Comment: A gender perspective for understanding low fertility in posttransitional societies”. In.: R. A. Bulatao, J. B. Casterline (eds.) Global fertility transition, supplement to Population and Development Review 27, New York, Population Council, pp. 177-183.

Pritchett, Lant H. 1994. “Desired Fertility and the Impact of Population Policies.” Population and Development Review. March, 20:1, pp. 1–55.

Reher, S. D. 1998. “Family ties in Western Europe: Persistent contrasts”. Population and Development Review 24 (2): 203-234.

Rendall, M. S. and S. Smallwood. 2003. “Higher qualifications, first-birth timing, and further childbearing in Englands and Wales”. Population Trends 111 (Spring 2003): 18-26.

Richerson, Peter J. and Robert Boyd, 2001, "Built For Speed, Not For Comfort, Darwinian Theory and Human Culture" History and Philosophy of the Life Sciences 23: 423-463 (2001) Special Issue Darwinian Evolution Across the Disciplines.

Riley, J. 2001. Rising Life Expectancy: A Global History. Cambridge: Cambridge University Press.

Rindfuss, Ronald R, Minja Kim Choe, Larry L. Bumpass, and Noriko O. Tsuya, 2004, Social Networks and Family Change in Japan , American Sociological Review 69 (December 2004): 838-861. Washington, DC: American Sociological Association.

Rindfuss, R. R., S. P. Morgan, and G. Swicegood. 1988. First births in America. Changes in the timing of parenthood. University of California Press, Berkeley, California.

Rostow, W.W. (1990) The stages of economic growth : a non-communist manifesto, 3. ed., Cambridge.

Sanderson, Warren C. and Sergei Scherbov, 2005. Average remaining lifetimes can increase as human populations age, Nature 435, 811-813

Santow, G. and M. Bracher. 2001. “Deferment of first birth and fluctuating fertility in Sweden”. European Journal of Population 17: 343-363.

Schelkle, W., Krauth, W-H., Kohli, M., and Elwert, G. 2000. Paradigms of Social Change: Modernization, Development, Transformation, Evolution, Frankfurt/New York: Campus/St Martin's Press

Schizzerotto A., Lucchini, M. (2002), Transitions to Adulthood during the Twentieth Century. A comparison of Great Britain, Italy and Sweden, EPAG Working Paper 2002-36. Colchester: University of Essex.

Schultz, Paul. 1994. “Human Capital, Family Planning, and Their Effects on Population

Growth.” American Economic Review. May, 84:2, pp. 255– 60.

Simó C., Golsh K. and Steinhage N. (2002), Increasing uncertainty in the Spanish labour market and entry into parenthood, GENUS, LVIII, 1, 77-119.

Sleebos, J. E. (2003), Low Fertility Rates in OECD Countries: Facts and Policy Responses, OECD Social, Employment And Migration Working Papers, OECD, Paris

Sobotka, T. 2004, Postponement of childbearing and low fertility in Europe. Doctoral thesis, University of Groningen. Amsterdam: Dutch University Press.

Sobotka, T. 2002, "Ten years of rapid fertility changes in the European post-communist countries". Evidence and interpretation, University of Groningen, Working Papers, 02-1

Sobotka, Thomas, Wolfgang Lutz and Dimiter Philipov, 2005, ‘Missing Births’: Decomposing the Declining Number of Births in Europe into Tempo, Quantum, and Age Structure Effects, European Demographic Research Papers, 2, Vienna Institute For Demography.

Söderling, I. 2005. Aiming for balanced population development. Presentation at the National Research and Development Center for Welfare and Health (STAKES), held in Helsinki, Finland, May 29, 2005.

Toulemon, L. and H. Leridon. 1999. “La famille idéale: Combien d’enfants, à quel âge?” INSEE Premiere No.652, June 1999.

United Nations, 2005, Human Development Report 2005, Oxford, Oxford University Press.

United Nations, 2000. Below replacement fertility. Population Bulletin of the United Nations, Special Issue Nos. 40-41. UN, New York.

United Nations Population Division. 2002. World Population Prospects: The 2002 Revision. New York: UN Department of Economic and Social Affairs.

Van de Kaa, Dirk J. 2001. Postmodern fertility preferences: from changing value orientation to new behaviour, Global fertility transition, Supplement to Population and Development Review, Vol.27, pp.290-331

Van de Kaa, Dirk. 1998. “Postmodern Fertility Preferences: From Changing Value Orientation to New Behavior.” Working Papers in Demography, No. 74. Canberra:

The Australian National University, Research School of Social Sciences.

Van de Kaa, D. J. 1994. “The second demographic transition revisited: Theories and expectations”. In.: G. Beets et al. (eds.) Population and family in the Low Countries 1993: Late fertility and other current issues. NIDI/CBGS Publication, No. 30, Swets and Zeitlinger, Berwyn, Pennsylvania/Amsterdam, pp. 81-126.

Van de Kaa, Dirk. 1987. “Europe’s Second Demographic Transition.” Population Bulletin. 42:1, pp. 1–57.

Van Nimwegen, N., M. Blommesteijn, H. Moors, and G. Beets. 2002. “Late motherhood in the Netherlands: Current trends, attitudes and policies”. Genus 28 (2): 9-34.

Weil, D. 2001, “Demographic Shocks: The View from History. Discussion”, in: Seismic Shifts: The Economic Impact of Demographic Change. Proceedings from the Federal Reserve Bank of Boston Conference Series No. 46.

Wetztel C. (2001), Squeezing Birth into Working Life: household panel data analysis comparing Germany, Great Britain, Sweden and the Netherlands, Aldershot, Ashgate.

Willekens, F. J. 1999. “The life course: Models and analysis”. In.: L. J. G. van Wissen and P. Dykstra (eds.) Population issues. An interdisciplinary focus. Kluwer Academic / Plenum Publishers, New York, pp. 23-51.

Willekens, F. J. 1991. “Understanding the interdependence between parallel careers”, In.: J. J. Siegers, J. de Jong-Gierveld, and E. van Imhoff (eds.) Female labour market behaviour and fertility. A rational choice approach. Berlin: Springer Verlag, pp. 11-31.

Wilmoth, John R. and Shiro Horiuchi, 1999. Rectangularization Revisited: Variability of Age at Death within Human Populations, Demography, Vol. 36, No. 4 (Nov., 1999) , pp. 475-495

Wilson, C. 2001. “On the scale of global demographic convergence 1950–2000,” Population and Development Review 27(1): 155–172.

Zapf, W. 1991. The Role of Innovations in Modernisation Theory, International Review of Sociology, New Series (3), 83-94

Four Stages Of The Demographic Transition

2008-02-11T12:49:00.000-08:00

Bo Malmberg's Theory Of Stages, Sanderson and Scherbov's Life Cycle Rescaling and Demographic Maturity

Transition in Phases?

The demographic transition, as is well known, has substantial effects on the age structures of populations (Lee, 2003a). The consequences of such age structure effects have been extensively analysed, and the extent of their impact on both the developed and the devolping world has been widely debated. While the literature on societal ageing contains numerous examples of empirical research documenting these changes (references), there have been surprisingly few attempts to evaluate the process on a theoretical level.

One notable exception in this context has been the Swedish demographer Bo Malmberg, who has proposed a four-phase typology of the demographic transition. This typology may be considered a useful starting point for examining the ways in which systematic age structure changes form a integral part of the transition process itself. In devloping his typology Malmberg takes as his core 'stylised fact' the steady (and seemingly unceasing) upward drift in median population ages which characterise the evolution of most transition and post transition societies following the initial fall in median ages which is associated with the sharp decline in child and infant mortality which characterises the first phase of the transition. What is perhaps most noteworthy and most interesting about Malmberg's work is his attempt to correlate such median age changes with the ongoing processes of social and economic evolution which accompany the 'long duree' of the demographic transition itself (Malmberg and Sommestad, 2000).

The modern demographic transition has, of course, had a substantial influence on both population-size and population growth-rates since its initiation in the United Kingdom at the end of the eighteenth century, and it is probably this aspect of the transition more than any other which has tended to capture the popular attention, at least, that is, until low fertility and population ageing started to hit the headlines. Less well appreciated and less well publicised, however, was the fact that the impact of the transition on the age structure of populations has been equally strong and significant. It is this age-structure dimension which Bo Malmberg, perhaps more than any other demographer, has thought about and has attempted to capture theoretically.

Impacts on age structure tend to be more extended in time than their more dramatic size-impact equivalents, and indeed one might claim that the whole process of the demographic transition in-and-of itself is best thought of as an extended and continuous process of age-transition. This is Malmberg's principle insight.

According to Malmberg, it is possible to break this age transition down into four distinct phases, each of which is characterised by the dominance of one specific age group (the term 'hegemony' does immediately come to mind here). With the onset of the transition (which is, remember, characterised by a sharp decline in mortality, and in particular in infant and child mortality) a child-dominated phase, comes into existence. It seems - in the words of one evocative metaphor - to be 'raining children'. Later, as the fertility transition itself takes hold and fertility begins its decline, there comes a young-adult stage. The acceleration of the fertility decline towards replacement fertility then produces what might be termed the middle-age - or prime adult worker - phase, and finally sustained below-replacement fertility produces in its wake an elderly - or mature - society.

Now before we proceed further, it is perhaps worth taking the time-out to examine an associated but exceedingly important issue: just how 'old' is old? This topic is an interesting one, since it conditions just how we read and interpret Malmberg's stages.

To help us with this Warren Sanderson & Sergei Scherbov have published a very interesting (and thought provoking) article in Nature (Sanderson and Scherbov, 2005), which explicitly addresses the issue of the relative value of the meaning of the word 'old' . The title of the article really tells the story in itself: "Average Remaining Lifetimes Can Increase as Human Populations Age". Put another way: we may now be facing the interesting enigma that, as societies - and collectively speaking - the longer we live, the longer we have left to live.

Riddles aside, what Sanderson and Scherbov actually propose is the establishment of a new metric: the 'standardised median age of the population' which is simply the crude median age figure standardised for expected remaining years of life. Now why would such a metric be interesting?

Well for one reason it would be interesting since it offers the possibility of better using median age as a rough and ready - rule of thumb type - guide for looking at all kinds of issues, from savings and investment behaviour, to social and political phenomena.

Effectively Malmberg suggests that societies with a median age of 'x' "may be expected to demonstrate the following characteristics..........". This approach has its attractions, since it is obviously on some trivial level true. Societies with a low median age and a lot of children need a plentiful supply of schools, teachers, and childminders, those with a high median age need rather old people’s homes, geriatricians and elderly care workers. But generalised too far it also would seem to have its drawbacks. I have long been aware that there is something intrinsically problematic about this way of doing things, and part of the source of my unease has been the gnawing feeling that median age is, as Sanderson and Scherbov explain, a moveable feast.

To try to get to the heart of the issue here, let’s take a simple analogy from economics. Economists often want to make comparisons, say about peoples relative preferences for apples and pears, and how these evolve through time, and in order to make such comparisons they often need to construct elaborate indexes like, for example, the consumer price index. Sometimes, however, economists just want to think about apples, and compare them, for example from one year’s harvest to the next, or from one variety to another. Now one simple way of making comparisons of this type is to think in terms of price, how this changes from one year to the next. But for many purposes this is a rather crude way of doing things, and often it simply isn’t sufficient, since a procedure of this kind ignores changes in quality. So to try and get round this kind of difficulty economists have invented (well better-put Zvi Griliches invented, see Griliches, 1967) a methodology known as hedonics. Hedonics is a procedure which allows us to make comparisons of the price and the quality of one and the same product across time.

Now, in a nutshell, this is what Sanderson and Scherbov are proposing to do for median age: formulate a hedonics of human age value, using remaining life expectancy as a crude proxy. As they say:

”Population ageing differs from the ageing of an individual. People who survive grow older with each year they live. Populations, on the other hand, can grow younger. Because a wide variety of matters such as the cost of medical care, retirement, bequests, consumption and the accumulation of human and tangible capital depend not only on age but also on time left to live, our understanding of population ageing must also re?ect both of these factors. Because conventionally measured old-age dependency ratios (the ratio of the number of people at the retirement age and above divided by the number of people in the working ages) have caused worry about the sustainability of pensions, it is important to recognize that these ratios, rescaled for life expectancy increases, are forecasted to change comparatively little over the century, suggesting caution in our assessment of long-term pension problems.”

One idea that Sanderson and Scherbov implicitly advance (although they don’t in my view make this sufficiently explicit) is that this 'getting younger' constitutes an effective re-evaluation of the prime age working life of the human individual. I think this is the point they are making when they say:

”Medical care expenditures provide an example where calculating the median remaining life expectancy in a population is useful. Health care costs rise rapidly in the last years of a person’s life. The change in the median remaining life expectancy between years is equal to the change in the median time to the onset of that phase of rapidly rising costs.”

That is to say, we shift the more feeble and fragile years up through the age course. In similar vein they also advance the idea of 'proportional life cycle rescaling', which effectively amounts to the same thing.

”Proportional life cycle rescaling is a heuristic not a predictive concept. It provides one simple way of thinking about a complex future in which the lengths of life cycle phases will be influenced by social policies and demographic constraints not modelled here. We use proportional life cycle rescaling by adjusting the conventional start of the working age phase (assumed to be age 20 in the year 2000) and the conventional end of that phase (assumed to be age 65 in 2000) proportionally to changes in life expectancy from 2000 onward.”

So the basic idea that we can take from Sanderson and Scherbov ( and it is an extremely important one) is that we need a benchmark (a calibration date, the year 2000, for example) and a moving-age-index calibrated in terms of the benchmark values. This would then give us a rough and ready but very 'user friendly' measure of just how the productive potential of a given population was changing through time (and, thrown in for good measure, some sort of indication of the future growth potential of the economy in question).

Their point about the relative lengths of the life cycle phases being, in part, socially constituted and thus influenced by social policies is not one to be taken lightly since it has two clear implications. Firstly, the necessary life course rescaling means a continuously later average start date for work. This is also what is meant by the idea of an economy moving continuously up the value chain, as the additional value can normally best be thought of as an increase in the ‘ideas content’ of the final product, and increasing ideas content means increasing education, and this continuing increase in education implies the age at which the young person may be considered fully productive for the new activities which are continuously so created simply rises and rises, as does the age of two other key life course events, setting up an independent home and having the first child.

Secondly it implies an upwardly flexible retirement age. This, as I say may be affected by both social realities (we remain more productive longer) and demographic constraints (rapid population ageing may mean that retirement ages need to rise simply to make the elderly dependency ratio economically sustainable). That there is nothing automatic about the occurrence of either of these ‘transitions’ can be seen from the debates about delayed retirement and other social policy and labour market reforms that are currently taking place across the European Union. Clearly a society's institutional structure is an important mediating framework between individual decisions and aggregate outcomes and there is nothing automatic at all about the evolution of this institutional structure (it is, as they say path dependent), a point which we shall return to later.

Now the question inevitably arises, what exactly has all this got to do with Bo Malmberg? Well Malmberg, let us remember, proposes a transition with a phases typology of age groups, and in thinking about this typology it is important to be aware, right from the outset, that these age groups are not carved in stone.

In the social and natural sciences three theoretical models have been proposed to help us understand human life events, Life History Theory (Biology/anthropology, Kaplan and Gangestead 2004, Charnov, 1993), Life Course Theory (sociology, anthropology, psychology, Mayer, 2001) and Life Cycle Theory (economics, Deaton, 2005).

Malmberg's proposal of age-structure phases (perhaps without his being aware of the fact) cuts clean across all three of these blocks of theory, and it does so in a way which can only be adequately understood in the light of the Sanderson and Scherbov proposal, since if "Average Remaining Lifetimes Can Increase as Human Populations Age" this increase must influence how we classify all the stages in the human life course, and, as a consequence, all the phases in Malmberg's typology. Put in other words, we need to apply the idea of proportional life cycle rescaling not simply to the phenotype, but to the genotype as well (as Sanderson and Schervov argue, population ageing differs from the ageing of an individual) so if we accept the validity of Malmberg’s initial project (and that for many, of course, is a big if) then we also need a 'four-phase transition' rescaling to accompany the individual life course one.

The Four Phases Themselves

Following Malmberg's initial typology, the first phase of the age transition, the child phase, occurs when falling infant and child mortality rates (coupled with improved pregnancy-delivery rates) produce an increase in the number of surving children in a given society. The reason this increase occurs is the fairly self-explanatory one that in earlier high-mortality populations, most of those who die are infants and children. In addition to this initial change, the subsequent process of cohort maturation means that the increasing numbers of children who survive into adulthood reach reproductive age and produce yet more children and so on, since the available fertile population grows continuosly, and this, in the absence of any compensatory downwards movement in fertlity levels, inevitably leads to further increases in the number of children born.

As a consequence at some stage after the initial fall in mortality (and the length of the intervening period involved varies from one country and one culture to another), the age structure of the population begins to assume the concave pyramid shape which has now become so familiar in the context of countries experiencing high rates of natural population growth.

Empirical experience has taught us that with the passage of time fertility rates generally start to gradually adjust downwards, as the demographic transition properly speaking sets-in, and under the impact of this fertility decline the pyramid structure starts to change significantly. Again there is nothing automatic about this process, and much of the research which was carried out under the aegis of the now legendary European Fertility Project was devoted to trying to establish precisely why it isn't, or put another way, what the factors are which influence the onset timing and subsequent intensity of the transition, or, using more contemporary language, why some countries 'stall' and others don't. (Coale and Watkins, 1986).

The reduction in the total number of children born means that the generations of new-born children are no longer – in pyramidal terms - as thick as they once were. More time passes, and the fertility reduction continues with fertility eventually dropping first towards, and then beyond, population replacement level. The steady reduction in cohort size eventually produces a bulge in the age structure - a bulge which pivots around those cohorts born just prior to the absolute decline in the number of live births. The classic population pyramid then changes its appearance into something else: as the base gets narrower the pyramid steadily assumes an increasingly convex shape (that this process is not always and everywhere the same can be seem by the exitence of the boom cohort phenomenon, especially noteworthy in its European and United states post World War II version, and in the impact that large and sustained immigration can have on the pyramid, once again the experience of the United States stands out here).

In the standard (or classical) model of the demographic transition the fall in the birth rate towards replacement level should constitute the last and closing phase of the transition (Lee, 2003). However, as we can now see, replacement fertility does not constitute an end point since the decline continues well beyond this level, with fertility moving onwards and downwards through replacement level, often reaching what some have called 'lowest-low' levels of fertility (1.2 or 1.3 TFRs, even though there is nothing self evidently 'lowest' about these levels in and of themselves, Kohler et al, 2002). However even in the absence of this continuing fertility decline the age transition process would be far from completed at replacement fertility level since a population which has developed a bulge in its age structure will be continuously transformed for as long as the bulge continues its passage across the various ages (with the baby-boomer phenomenon in the United States being currently a good example of just this situation).

Now, armed with this fourfold typology as their starting point and with Swedish data from the late 18th century onwards as their raw material, Malmberg and Sommestad (2000) have tried to specify some of the ideal-type phases (using an expression of Max Weber's, which, it should be noted, Malmberg and Sommestad do not adopt) of the age transition as follows:

The first phase - the child one - began in Sweden in the early 19th century and continued up to the middle of the century. The second phase, the young adult one, started around 1840 and continues to around the time of the first world war. The third phase, that of an expanding middle aged population, started in the 1870s and continued up to the 1980s. Meanwhile the fourth phase, with an expanding old age population, really started in the 1980s and is of course still ongoing. (A word of caution needs to be expressed here since Malmberg's dating is somewhat vague, and seems to incorporate the idea of overlapping phases, a matter which complicates the problem enormously methodologically speaking. Nonetheless, this issue will be sidestepped here since the intention is not to rigourously identify the stages, but to justify them theoretically, and to provide a more rigourous underpinning and interpretation of the whole idea).

Well, so much for the initial theory. But isn't it somewhat crude? Doesn't it simply replace primitive economic determinism with another demographic version which is ridden with all the same attendant problems? I would argue that it does not. In the first place the theory is not really a theory at all, but simply a description: it merely states that during the roll-out of the modern demographic transition it is possible to note a certain correlation between various types of social and institutional behaviour and an evolving age structure. In particular there is no ascription of directional causal arrows here.

Applied rigidly this would obviously immediately become open to all the old criticisms which were initially applied to the most schematic versions of the original transition theory. If, on the other hand, Malmberg's taxonomy is simply treated as a generalisation with significant descriptive properties it seems useful and remarkably insightful.

Now, as as been indicated, one way of taking Malmberg's initial idea forward is to follow the path marked out by Sanderson and Scherbov. However, another useful strategy might be to try and conceptualise the idea in terms of the various life event theories which have been advanced from within the social sciences: Life Cycle Theory (economics), Life Course Theory (sociology) and Life History Theory (biology/anthropology).

Life Course Theory

Sociologists tend to find life course theory useful since it helps them conceptualise the ways in which patterns in life courses serve as mechanisms through which social structures are generated as the aggregate outcome of a totality of individual decisions taken across the life course. One good and highly relevant example here would be the way in which the age and cohort structure of a population is the result of a multitude of individual fertility behaviours and decisions. Looked at the other way round, individual behaviour is itself conditioned by a variety of social and economic constraints, constraints which determine the options available for each individual actor. In this sense any micro-level behaviour is influenced by its given context - this defining feature of human agency has often been referred to by sociologists as "embeddedness", which is a multidimensional construct relating generally to the importance of social systems and networks for actors, an idea which draws attention to the fact that actors integrated within multiple social and institutional networks face varying resource and constraint sets according to the problem at hand (Granovetter, 1985, Luhmann, 1995)

In similar fashion, the decisions concerning major life course events are shaped by an individual’s past experiences and life events. This is sometimes referred to as being the outcome of a process of life contingency and cumulative causation. As Dykstra and van Wissen assert, “early events pave the way for some roles and preclude others, altering life chances and prospects” (Dykstra and van Wissen, 1999: 12). Individuals here are seen as rational actors trying to utilise the information available to them to achieve their assumed goals within a variety of constraints, be they material resources, institutional infrastructures, or cultural norms.

In life course theory, notions of rational orientation and of the interaction between competing itineries are also linked with the concept of strategic behaviour, where the timing of life structuring decisions like parenthood may be conceptualised as forming part of a strategy through which agents attempt to organise their life course. In the context of parenting decisions, much of the work in this area has particularly drawn attention to problems arising in attempts to coordinate employment and family life (Brewster and Rindfuss 2000).

On this level life courses may be considered not simply as the life histories of persons as individuals but as patterned dynamic expressions of social structure. Such dynamics operate in populations or subsets of populations. They are governed intentionally or unintentionally by institutions, and are the intentional or unintentional outcomes of the behavior of actors.

One standard theme which may be found among most life course theorists is that of the diminishing contemporary importance of the ‘standard’ life biographies that seem to have characterised the industrial age. This 'destandardisation process' seems to have been gathering steam in the highly industrialised countries since the late 1960s (Mayer 2001) and side-by-side with this we have seen a closely associated shift towards less prescriptive and more variable life biographies, or 'choice biographies' as some have called them. According to Lesthaeghe “life cycle transitions have become more frequent, less strictly patterned, and more complex.” (Lesthaeghe, 1995: p18) , and this has produced more individual diversity in the timing of various life course transitions (Heinz and Krüger 2001).

At the heart of this recent structural transformation in the life course lies the changing character of the young adult years. These years constitute what some have termed a ‘demographically dense’ period, in that more life-course events occur between the ages of 18 and 29 than at any other stage of the life course (Rindfuss 1991).

During the last quarter of the twentieth century the lives of young adults living in 'post industrial' societies have changed dramatically. This change can be seen in a whole variety of settings: in prolonged education, evolving and flexible social norms, changing patterns of home leaving and partnership formation, and in the arrival of a whole new card deck of risks and uncertainties. The term 'destandardisation process' is now used by many to describe the way the “classic sequence of finishing school, entry into labour force, home leaving linked to marriage and subsequent parenthood is being reordered in ever larger segments of the population.” (Lesthaeghe and Moors , 2000: p153)

The proces here is a complex one since, on the one hand, young people today enjoy an unprecedented freedom of choice in personal matters (with many of them embracing what Mayer has loosely termed ‘hedonistic individualism') wherein young adults assume responsibility for their own “life designs and life projects or, rather, follow egoistically the shifting material incentives and consumption idols from situation to situation” (Mayer, 2001).

On the other hand, new constraints and new uncertainties, ones which are related above all to employment stability, have emerged in the lives of young adults. In particular young adults are increasingly having to cope with the existence in perpetuity of what would have previously been considered to be non-standard employment contracts, where spells of unemployment are regularly interspersed with periods of part-time work and part-time study and, from time to time, full-time but short-duration employment contracts. For females the situation has become even more complex, since they often find themselves “alternating between non-standard forms of participation in work, education, and family life”.

In general, the life-transitions that traditionally served as the milestone indicators for reaching adulthood - such as leaving the parental home, marrying, and entering parenthood - have both moved upwards in terms of age, and become more flexible as to their timing, and they are now certainly largely to be found outside the long-accepted boundaries which once separated youth and adulthood. And all of this has been accompanied by a growing fluidity and diversity in partnerships and pathways to parenthood (Rindfuss 1991). So rather than a strict ‘sequencing' process, what we now see is the operation of a number of commonly shared concepts of what constitute the necessary preconditions for parenthood in operation. Foremost among these are: leaving the parental home, finishing education, and accumulating resources.

The foundation block for the entire process is really economic independence, and it is the interpretation of what this means which governs the household formation process. One implication of this in an era of heightened individualism and rising consumer aspirations is the idea that the household needs to depend on two permanent incomes rather than one and this in itself may delay marriage and childbearing, and particularly in the more 'familiaristic' societies of Southern Europe (see Dalla Zuanna, 2001 for the Italian case).

Males in particular, it has been argued, often attach considerable importance to the achievement of certain classes of consumer goods, and their appetite for, for examply, new-tech products may well get priority over family formation (Goldscheider and Kaufman 1996: 89). Females on the other hand may attach more importance to the partnership selection procedure given the growing uncertainty about partnership stability and the high level of parental investment required of both partners in raising fewer but better prepared children. The increase in both own- and offspring- investments which the quality/quantity trade-off in offspring implies has lead a number of researchers to speculate that this process may well lead to greater selectivity on the part of females when identifying a suitable partner with whom to co-invest in children. This selectivity is on the one hand a result of the crucial role of both maternal and paternal skills-based investments in the development of offspring, and the unequal distribution of these investments between the sexes. (Kaplan, 2003)

Summarising all this, Mayer indicates that the principal components of the post-transition life course regimes are:

• the age of leaving the parental home,

• the age of leaving school or formalized training,

• the process of labor market entry,

• the rate of fluctuation as work life mobility between firms,

• the rate of work life mobility between occupations,

• the shape and distribution of income trajectories,

• the degree of career involvement of women,

• the fertility and the stability of partnership unions,

• the median retirement age and its degree of dispersion

(Mayer, 2004)

One major structural feature of all life courses is evidently their internal temporal ordering, that is, the relative duration of each of the various states which the life course embodies, as well as the age distributions of the various events or transitions.

Using data from the West German part of the German Life History Study Brückner and Mayer have illustrated just how the timing of key life events such as age of school leaving, termination of education, leaving home, first marriage and age at first childbirth varied among German men and women across the twentieth century (Brückner and Mayer, 2005). Interestingly the educational data reveal a constant upward movement in termination dates while the family decisions timing is more complex, first falling and then once more rising.

Median age of school leaving for males, for example, rises from 14.3 for the 1920 birth cohort to 17.2 for the 1971 one (for females the figures are 14.4 in 1920 and 17.6 in 1971). Median age at first employment for males rises from 18.1 for the 1920 cohort to 20.3 for the 1971 one (for females the figures are 16.9 in the 1920 case and 20.4 in the 1971 one).

Median age on leaving home first falls, before then once more rising, for males falling from 29.7 for the 1920 cohort to 23.3 for the 1950 one and then rising to 24.2 in the 1971 cohort (for females the figures are 28.3 in 1920, 20.8 in the 1955 case and 21.8 in the 1971 one).

Median age at first marriage reveals a similar pattern, with the age for males falling from 27.7 for the 1920 cohort to 25.8 for the 1950 one and then rising to 29.2 in the 1964 cohort (for females the figures are 23.3 in 1920 , 21.5 in the 1950 case and 25.5 in the 1964 one). Similarly with median age at first childbirth, the age for males falls from 29.7 for the 1920 cohort to 27.3 for the 1940 one and then rises to 32.6 in the 1964 cohort (for females the equivalent figures are 25.4 in 1920 , 23.8 in the 1940 birth cohort case and 28.1 in the 1964 one).

One perennial question which arises in this context is just how does such ordering and regularity in life courses come about?

The sociologist Karl Ulrich Mayer has given three answers to this question (Mayer, 2004). Ordering and regularity arise due to:

1/. the process of internal differentiation of societies into subsystems or institutional fields

2/. the internal dynamic of individual lives in group contexts.

3/. the fact that the ordering is not simply the result of the interaction of society on the one hand and the individual on the other but that there is an intermediary structure of aggregates of individuals in the form of populations such as birth cohorts or labor market entry cohorts and the mediation of the activity of these aggregates through institutions.

An evident question then arises: what were the institutional configurations that characterised the various life course and age structure regimes that have been identified by Malmberg and Sommestad above? The following may be considered a preliminary suggestive list:

In traditional societies the life course regime was regulated by the demographics of high mortality and high fertility, and by the prerequisites and vicissitudes of a rural economy which was organised in the absence of such stabilising elements as the agrochemical fertilization or the use of scientific animal husbandry. Following the onset of the transition existence became fragile and children were many. This would be Malmberg’s child-dominated society. The structure of life course decisions reflects this instability, in particular this can be seen in the strategic decision of females to generally reproduce at an early age.

The early industrial life course regime is formed of course against the background of an incipient capitalist economy, in the presence of comparatively weak and unstable labour movements and - under the impact of the early stages of the modern demographic transition - a high labour supply. This is Malmbergs young adult society, with continuing large (if reducing) birth cohorts and weak political, legal and institutional infrastructure. One consequence of this was and is the widespread existence of large scale emigration.

The life course regime in the later industrial societies normally occurs in the presence of an implicit contract between capital and labour, often accompanied by the existence of mass membership trade union organisations. The industrial society was characterised by the presence of mass production technologies and mass 'standardised' consumption goods. Macroeconomic policy intervention increasingly stabilized economic cycles, and full life-time employment, rising real wages and standards of living, and welfare state insurance regimes characterise the social and economic sphere. In other words the regime evolves in the presence of a strong and 'unifying' institutional framework which includes sub-systems (like the educational one) which nourish standardisation. This would be Malmberg’s mature adult society.

With the arrival of the postindustrial, or knowledge, or services-dominated societies, the life course regime has once more witnessed significant changes: sustained educational expansion into the tertiary stage, growing female emancipation, accelerating value changes, growing individualization and self-direction, weakening of organised labour, growing de-industrialization, flexibilised labour markets and the ending of life-time employment, globalization of economic markets, and sizeable demographic changes (fuelled in particular by an extended postponment of childbearing), low fertility levels and greatly increased life expectancy the combined operation of all of which has lead to a significant inversion in the classical age pyramid, and, of course, the arrival of Malmberg’s elderly society. This process has been accompanied by a loosening of the late-industrial institutional framework with a continuing degregularisation of the welfare system, a weakening of 'standard' life biographies, and the flexibilisation of nearly everything.

One way in which many life course theorists have treated these transitions has been to focus on changes in what has become known as the 'welfare mix' (i.e., the relative importance and manner of interconnectedness of economic markets, the family, and the state: see Esping-Andersen, 1999) and the age-directional flow of resources (Lee, 2003b).

Applying the Theory to Sweden

Now if we attempt to look at how these two sets of ideas (life course theory and proportional life cycle rescaling) might be applied to Malmberg's original template , the first thing which should strike us is that different regions in the world have, naturally, experienced the demographic transition (as well as the accompanying age structure and life course transitions) at different points in time, and with varying degrees of intensity. However we could equally note that once the transition is actually under way, the different regions tend to pass through Malmbergs four phases in remarkably similar fashion.

The most important consequence of the age structural changes, at least when viewed from a life cycle perspective, is the fact that an individual's productive capacity and their economic behaviour vary across the cycle. Newborn humans are unable to survive without the adult support. Many years of care, education and training are needed before children have acquired the cognitive skills and productive potential of an average adult.

Similarly, once we have aged beyond a certain point (even though, following Sanderson and Scherbov's proportional life rescaling idea, this point may constitute a kind of moving target), our productive capacity will inevitably begin to decline, until finally our ability to fend for ourselves falls short of what we need to provide for our own survival.

In sharp contrast to both these stages of the life cycle, the years in-between are a time when most adults have a capacity to produce which goes well beyond what they need to guarantee their own immediate survival, and indeed in post-industrial societies seem able to furnish the wherewithall to satisfy what appear to many to be even the most exotic of needs. Put simply, our productive capacity initially is limited, grows steadily with age, rises to a peak and then eventually enters decline, in the course of this life cycle the hump-shaped curve traditionally associated with lifelong earnings and productivity is produced. Essentially the curve is the product of an age related trade-off between experience and speed. (Richerson and Boyd, 2001)

The Life Course/Malmberg Fushion

Using Weberian ideal-type terminology, Malmberg suggests that child dominated societies tend to be inherently politically unstable. They also tend to be poor. A statistical analysis carried out by Malmberg and Sommestad shows that a one percent increase in the share of children in a society is associated with a 2.5 percent increase in the poverty rate. Indeed they find that the share of children in a population explains slightly more than 50 percent of the regional variations in poverty. The sub-Saharan African states have at present remarkably high child dependency rates, and these countries are of course also among the poorest in the world. (Bloom and Sachs, 1998).

Another characteristic feature of child abundant economies is the widespread presence of child labour, a phenomenon clearly connected to their state of poverty. (Cunningham and Viazzo 1996, de Coninck-Smith et al 1997).

A third characteristic feature of child abundant countries is a strong dependence on the exploitation of natural resources. In Sweden, for example, the area of land under cultivation roughly doubled during the early phase of the age transition, from 1820 to 1865. Such cultivation reached a maximum around 1920, a point in time when the proportion of children in the population was at its height. From this time on expansion in agriculture became intensive rather than extensive, with productivity improvements meaning that increased output could continually be achieved with the cultivation of ever smaller total quantities of land.

Today, land reclamation is still an important issue in many countries with rapidly-growing, child-abundant populations. Countries, for example, in sub-Saharan Africa. Between 1980 and 1996 land reclamation was particularly intense in West Africa, with increases between 20 and 40 percent in the share of arable land in countries like Gabon, Guinea, Burkina Faso, Ghana and Côte d'Ivoire.

Finally, child abundant countries are enormously dependent on foreign capital. Malmberg directs us to the work of Lennart Schön, who has systematically studied the history of Swedish trade and the behaviour of its current account balance. He found that Sweden’s current account status was negative for 60 years, from the 1850s up to the end of the first decade of the twentieth century . Indeed, by 1910, Sweden was one of the most indebted nations on the planet, with total estimated debt amounting to some 75 percent of GNP (Schon, 1994).

In Sweden the young-adult phase (from 1840 to WWI) was strongly associated with modernisation (both political and social). This was a time of agricultural transformation, liberalisation, railway building, emigration, urbanisation, industrialisation, popular movements, and, towards the end of the phase, rapidly falling birth rates. New industries emerged, international trade developed, and financial markets boomed and collapsed.

Furthermore, the increase in the share of young people coincided with increasing social and political conflict, and, as a response to this, there was increasing democratisation and the development of a more extensive state involvement in the economy.

The third phase (the rapid expansion of the prime age worker population) was characterized in Sweden by the evolution of much more stable economic conditions. In Sweden this phase probably commenced in the 1930s and continued up to about 1980. If we look for common characteristics among countries that have entered this phase, the most obvious feature which stands out is the presence of sustained economic growth. Countries that for a number of decades have benefited from increases in the the specific weight of their prime age workers in most cases have gone on to become full-members of the OECD. An increase in the group of middle-aged people is thus clearly associated with a more developed stage of economic growth, a stage that the economist Walt Rostow once designated ”the drive to maturity” (Rostow, W.W. 1990)

The final phase of the age transition, the "mature/elderly" phase, is largely a late 20th century phenomenon. On the Malmberg schema this phase is reached when the enlarged cohorts which mark the earlier turning point reach retirement age. Concurrently fertility continues its decline to below replacement levels. As this phase is only really now opening it is hard to specify clearly the eventual defining characteristics.

Two features, however, do stand out.

First, ageing countries have experienced a decline in their rate of economic growth, along with a number of related negative economic trends (like an associated weakening in consumer demand and a rise in the level of saving in excess of demand for investment). Statistical analyses of OECD data reveal a negative association between, on the one hand, the population share aged 60 and above, and, on the other, per capita income growth, productivity growth, rate of capital formation in the business sector, and housing investments. (Analyses based on data in OECD, Statistical Compendium, Vol. 97:1, Paris, 1997; Lindh and Malmberg 1999)

Secondly, population ageing in the OECD countries has been closely connected with growing public expenditure and budget deficits. There is thus a strong positive association between, on the one hand, the share of people aged 60 and above, and, on the other, government consumption, public sector employment, public debt, and taxes paid. All in all, these correlations indicate that the elderly phase of the age transition is associated with a major structural shift in the economy away from a traditional high-growth society towards an economically less vigorous welfare-dependent state. If these results are compared to the life cycle behaviour story, it is evident that the decreasing productive capacity that marks individual ageing also translates itself into significant macroeconomic consequences.

Malmberg uses the terminology elderly society for this fourth stage, but there may well be good theoretical justification for abandoning this use. In particular the expression 'elderly society' implies in and of itself the end of some particular road. For that reason the term 'mature society' will here be preferred. Since this expression implies that something stable - like, for example, the homeostatic end state of a completed transition - has been achieved. In this sense it becomes something normal, and to be welcomeed, rather than something strange and preoccupying.

The terms 'ageing' and 'elderly' society do on the other hand have a certain validity when applied in the context of either an excessively rapid decline in fertility (or rise in life expectancy) which change the balance between the various (age rescaled) life course groups. Thus a rapid first birth postponement process may produce what some have termed a 'birth dearth' while it is taking place, and this, of course, can lead to excessive ageing which cannot be compensated for by changes in age-related capacities. Also institutional (path dependent) rigidities which impede fexible adaptation to the ongoing life-course and life history changes (such as, for example, failure to adapt the working age and retirement classification systems) can produce social and economic 'bottlenecks' which may, if left unchecked, give renewed meaning to that vastly overused expression 'the ageing problem'.

Conclusion

In a recent presentation, Lutz, Skirbekk and Testa (Lutz et al, 2005) argued the following:

"Viewed under a long-term perspective, the demographic transition taught us that the balance of births and deaths (if we assume that systems eventually will move towards homeostasis) can be disturbed

for many decades due to the fact that fertility is strongly embedded in the system of social norms and that such demographic regimes can be very persistent once they are well established."

"Because of this we have seen many decades of “overshooting” birth rates that have resulted in historically unprecedented population growth. It cannot be ruled out that the same forces of social momentum, once a new low fertility regime has been established, will result in decades of “undershooting” birth rates resulting in historically unprecedented population ageing".

Now this statement is interesting in the sense that it assumes that the balance of births and deaths may be regulated homeostatically, even if the homeostatic mechanism which does this is not specified. This allows the authors to use terminology like 'overshooting' and 'undershooting' in relation to fertility.

In this paper I have argued that thinking about this demographic transition process in the context of Bo Malmberg's idea of four phases for the transition and mapping Malmberg's insight onto Sanderson and Scherbov's idea of 'proportional life cycle rescaling' and via the template offered by life course theory might provide us with a very useful framework for understanding the transition process, in particular as it enables us to capture the transition as one from youth to maturity, in both the chronological age and in the institutional senses of the term.

Basically it has been argued here that the Malmberg idea is fruitful but flawed. It is flawed since since it is excessively mechanistic: it is almost a modern 'demographic materialism'. It also has the problem that simply speaking of an elderly society tout court leaves us scratching our heads trying to answer the rather problematic question of what happens next.

If however we map Malmberg's phases onto Sanderson and Scherbov life cycle rescaling we may

arrive at a more theoretically interesting destination: social maturity. If, instead of terming the last phase

in Malmberg's typology the 'elderly society', we simply called it the 'mature demographic regime' then, thinking homeostatically, and imagining for a moment that homeostasis might mean that declining fertility and increasing life expectancy just balance each other , while the proportional life cycle rescaling property works so as to just keep the balance in balance, then we may have what so many people have long been looking for: demographic sustainability.

Of course, as one rather well-known German-speaker once said 'theory is grey, while life is green', so things will never actually be just like this.

Howver this is where Lutz's 'overshooting' and 'undershooting', fertility traps, and path-dependent lock-ins etc etc etc come into play. The bottom line here is that while this way of looking at things would not solve all our problems with a click of the fingers, at least it would offer us a sound theoretical yardstick with which to assess the problems themselves.

References

Bloom, D.E., and J. D. Sachs. (1998)., ”Geography, Demography and Economic Growth in Africa”, Brooking Papers on Economic Activity 1998.

Brewster, K. L. and R. R. Rindfuss. 2000. “Fertility and women's employment in industrialized nations”. Annual Review of Sociology 26: 271-296.

Brückner, H. & Mayer, K. U. (2005). “De-Standardization of the Life Course: What It Might Mean? And If It Means Anything, Whether it Actually Took Place,” In R. Macmillan (Ed.), The Structure of the Life Course: Standardized? Individualized? Differentiated? (Vol. 9, pp. 27-54). Amsterdam et al.: JAI Elsevier.

Coale, Ansley J., and Susan C. Watkins, eds. (1986). The Decline of Fertility in Europe. Princeton, New Jersey: Princeton University Press.

Charnov, E. L. (1993). Life history invariants: Some explanations of symmetry in evolutionary ecology. Oxford: Oxford University Press.

H. Cunningham and P. P. Viazzo (eds.), Child Labour in Historical Perspective 1800-1985: Case Studies From Europe, Japan and Colombia, Unicef, 1996;

N. de Coninck-Smith, B. Sandin and E. Schrumpf (eds.), 1997, Industrious Children. Work and Childhood in the Nordic Countries 1850-1990, Odense.

Dalla Zuanna, G. 2001. “The banquet of Aeolus: A familistic interpretation of Italy’s lowest-low fertility”. Demographic Research 4, Article 5: 133-161. «www.demographic-research.org».

Deaton, A. 2005. Franco Modigliani and the Life Cycle Theory of Consumption, Paper Presented at the Convegno Internazionale Franco Modgliani, Accademia Nazionale dei Lincei, Rome, February 17th–18th, 2005. Banca Nazionale del Lavoro Quarterly Review, Forthcoming

Dykstra, P. and L. van Wissen. 1999. “Introduction: The life course approach as an interdisciplinary framework for population studies”. In.: In.: L. van Wissen and P. Dykstra (eds.) Population issues. An interdisciplinary focus. Kluwer Academic / Plenum Publishers, New York, pp. 1-22.

Esping-Andersen, G. 1999. Social foundations of postindustrial economies. Oxford University Press, Oxford.

Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Princeton, NJ: Princeton University Press.

Goldscheider, F. K. and G. Kaufman. 1996. “Fertility and commitment. Bringing men back in”. In.: J. B. Casterline, R. D. Lee, and K. A. Foote (eds.) Fertility in the United States. New patterns, new theories. Supplement to Population and Development Review 22, Population Council, New York, pp. 87-99.

Granovetter, M. S. 1985. "Economic Action and Social Structure: The Problem of Embeddedness."

American Journal of Sociology 91:481-510.

Griliches, Zvi. “Hedonic Price Indexes Revisited: Some Notes on the State of the Art, Proceedings of the Business and Economic Statistics Section, American Statistical Association, 1967, pp. 324-332.

Heinz, W. R. and H. Krüger. 2001. “Life course: Innovations and challenges for social research”. Current Sociology 49 (2): 29-45.

Kaplan, Hillard S. & Steven W. Gangestad. 2004. Life History and Evolutionary Psychology, University of New Mexico, mimeo

Kohler, H.-P., F. C. Billari, and J. A. Ortega. 2002. “The emergence of lowest-low fertility in Europe during the 1990s”. Population and Development Review 28 (4): 641-680.

Lee, Ronald D. 2003a. The Demographic Transition: Three Centuries of Fundamental Change. Journal of Economic Perspectives, 17 (Fall 2003), 167-190

Lee, Ronald D. 2003b. Demographic Change, Welfare, and Intergenerational Transfers: A Global Overview, Genus, v. LIX, No. 3-4, pp. 43-70, July-December 2003.

Lesthaeghe, R. 1995. “The second demographic transition in Western countries: An interpretation”. In.: K. O. Mason and A.-M. Jensen (eds.) Gender and family change in industrialized countries. Oxford, Clarendon Press, pp. 17-62.

Lesthaeghe, R and G. Moors. 2000. “Recent trends in fertility and household formation in the industrialized world”. Review of Population and Social Policy No. 9: 121–170.

Lindh, T. and B. Malmberg (1999). "Age structure effects and growth in the OECD, 1950-1990." Journal of Population Economics 12: 431-449.

Luhmann, N. 1995. Social Systems, Stanford University Press, Stanford

Lutz, W, Vegard Skirbekk and Maria Rita Testa. 2005. "The Low fertility trap hypothesis", presentation given at the conference "The Postponement of Childbearing in Europe", Vienna Institute of Demography, Vienna

Malmberg, Bo & Lena Sommestad. (2000). Four Phases in the Demographic Transition, Implications for Economic and Social Development in Sweden, 1820-2000. Arbetsrapport/Institutet för Framtidsstudier; Working Paper 2000:6

Mayer, K. U. (2004). Whose Lives? How History, Societies and Institutions Define and Shape Life Courses. Research in Human Development 1 (3), 161-187.

Mayer, K. U. 2001. “The paradox of global social change and national path dependencies: Life course patterns in advanced societies”. In.: A. E. Woodward and M. Kohli (eds.) Inclusions and exclusions in European societies. Routledge, London, pp. 89-110.

Richerson, Peter J. and Robert Boyd, 2001, "Built For Speed, Not For Comfort, Darwinian Theory and Human Culture" History and Philosophy of the Life Sciences 23: 423-463 (2001) Special Issue Darwinian Evolution Across the Disciplines.

Rindfuss, R. R. 1991. “The young adult years: Diversity, structural change, and fertility”. Demography 28 (4): 493-512.

Rostow, W.W. (1990) The stages of economic growth : a non-communist manifesto, 3. ed., Cambridge.

Sanderson, Warren C. and Sergei Scherbov, 2005. Average remaining lifetimes can increase as human populations age, Nature 435, 811-813

Schelkle, W., Krauth, W-H., Kohli, M., and Elwert, G. 2000. Paradigms of Social Change: Modernization, Development, Transformation, Evolution, Frankfurt/New York: Campus/St Martin's Press

Schön, Lennart. Omvandling och obalans. Appendix 3 to Långtidsutredningen 1994, Stockholm, 1994, p. 47-54.

Zapf, W. 1991. The Role of Innovations in Modernisation Theory, International Review of Sociology, New Series (3), 83-94

The Rediscovery of Age Structure

2008-02-11T01:01:00.000-08:00

The demographic question, the issue that is of just how important demographic processes actually are in the emergent evolution of economic and social phenomena, is one which has enjoyed a long and complex history. It has been one of the most hotly contested social science topics since it seems time immemorial, or at least this is how it feels. Protaganists of now this, now that point of view have each argued their point of view with equal fervour and conviction.

Back in the 18th century, for example, mercantilists and physiocrats held that having a large population was a good thing, something to be encouraged, since, in addition to enhancing national security, it was thought to be a major stimulus to economic growth. This argument has long enjoyed a certain popularity, and has continued to rear its head now here, now there across the span of the years in the form of either pro-natalist public policies or economic priorities which could possibly be best be described as 'immigration friendly'. On the other hand there have always been those who have seen in growing populations a threat and a menace. Thomas Malthus would be the name which first comes to mind here, but there have been plenty of others, whether these have been at the end of the 19th century (with the densely overcrowded cities of Western and Northern Europe) or in the second half of the twentieth century (with the population explosion which followed the mortality decline in the third world).

It is though, undoubtedly, the arrival of below replacement fertility in most of the societies which have achieved maturity in their economic systems which has lead produced the deepest, and possibly the most critical, episode in this long series of heart-searching debates. Actually the current debate is in many ways a revival of one which originated in the back in 1930s - when the prospect of generalised below replacement fertility first began to be glimpsed by the most far-sighted theorists . At the time writers and theorists began to forsee important negative (even cataclysmic) consequences - both socially and economically - for those countries who were expected to experience population decline. (Redaway, 1939, Myrdal, 1940, Spengler, 1938). Of course the decline never happened, war came and was followed by a generalised baby boom, and demographers, not for the first or for the last time, were felt to be hopelessly incapable of predicting anything. However, as we can now see, these early theorists were not wrong, but simply out of time.

In economic theory, the question of population change has had a complex history. Following the early lead of the mercantilists many have considered it to be a positive influence on economic performance due to its impact on aggregate demand and investment. In the context of development theory population growth has often been argued to stimulate development through its effect on the investment share in GNP.

This type of argument has come and gone over the years as it has come under contsant and sustained attack from neo-Malthusians who have argued that excessive population growth has been one of the decisive factors maintaining countries in povery. The origins of this argument can, of course, be traced back to the early 19th century, with Malthus himself arguing that population growth, by producing decreasing returns in agriculture, leads directly to lower per capita income.

This decreasing returns argument itself has in its turn been countered by those who have tried to suggest that in the industrialisation process - whether in the original industrial revolution in the UK, or in the modern 'development process' in third world countries - scale effects may well be important, but this time with the effects working in the opposite direction: ie the increasing returns on investment produced by growing domestic demand being thought to be one of the decisive 'trigger' elements in the take-off process (Kuznets, 1960).

More recently there have been those working within what could broadly speaking be termed the 'new growth theory' tradition who have argued that it is precisely the increasing returns effect produced by the non-rivaly properties of ideas which may be the best argument of all in favour of population growth (Kremer, 1993, Jones, 2001).

Over the years versions of both these classes of arguments have tended to come and go, with the pendulum swinging now this way, now that.

In general, neo-malthusians (like the Swede Knut Wicksell at the end of the 19th century) have continued to argue the view that population growth is harmful, while Keynesians, for example, have generally tended to see population growth as a stimulus for investment demand and, thus, for income growth (Perlman 1975). A third, more neutralist view, has, however, come to occupy the position of 'paradigmatic view' from the early 1970s onwards. This view holds that population growth rates are not an economically decisive factor, one way or the other, and that they do not constitute a significant variable when it comes to understanding differences in per capita income growth across countries. This is essentially the view taken in the standard accounts of neo-classical growth theory. (Mankiw et al, 1992, Jones, 2001, Barro and Sala i Martin, 2003).

This view could fairly be said to have constituted itself as the dominant academic consensus in the area of growth theory since the early 1980s, with the neutrality of population change for per-capita income being treated as being a theoretically well-founded result. This state of affairs has, of course, not been without its critics, with the claim being made that this finding has in and of itself had a negative impact on the development debate since it has lead to the marginalization of population and reproductive health questions as instruments of economic development policy within key agencies like the World Bank (Birdsall 2003, Kelley 2003).

The heyday of the neutralist view, however, was in many ways the 1990s, as throughout the decade growth study after growth study seemed to reveal little in the way of cross-country evidence to justify thinking there was any significant demographic effect on growth, either in the form of a dividend or in that of a penalty. However, it is important to stress right from the outset that much of the work which belonged to this generation of population 'neutralist' growth research had one obvious limitation: virtually every other factor which could conceivably infuence growth was held constant in order to test exclusively for correlations between rates of population growth and rates of per capita income growth.

Now whether results obtained in this way truly reflect the unimportance of population growth, or, as some have argued, they simply reveal the summative outcome of the impact of diffent, and mutually offsetting, negative and positive influences of population on economic growth is a question which is still, obviously, being debated.

Be this as it may, in recent years the 'neutralist' view has been coming under increasing criticism. On the technical level, critics of the neutralist studies tend to cite the existence of inadequate control variables or other such model specification errors, or the relatively poor quality of the available data, or the presence of reverse causality, or all of these combined.

At the same time, from the late 1980s onwards another tradition of growth research has increasingly come to cast doubt on the idea that population dynamics don't matter for economic theory, and it has done so by emphasising that it is not population growth per se which is important, but the impact of changes in the rate of population growth on the age structure of a population. This increased interest in age structure coincided, not co-incidentally, with the arrival of South Korea, Taiwan, Hong Kong and Singapore on the scene as newly developed economies, an arrival which has often been termed the Asian Tigers phenomenon.

The arrival of this change in attitudes can be traced to new evidence coming in three principal forms.

In the first place a series of empirical studies based on aggregate-level panel data have concluded that demographic factors do in fact have a strong, and statistically significant effect on aggregate saving rates (Bloom at al, 2003; Deaton and Paxson, 2000; Kelley and Schmidt, 1996; Kinugasa, 2004; Williamson, 2001) and on economic growth (Bloom et al, 2001; Bloom and Williamson, 1998; Kelley and Schmidt, 1995).

Secondly, detailed case studies of the East Asian miracle have provided compelling and consistent evidence that what has come to be known as the "demographic dividend" was an important element in facilitating the economic success of the countries concerned (Bloom and Williamson, 1998; Mason, 2001b; Mason et al, 1999). In particular, and in a study which really made it impossible to continue to simply ignore the relevance of demographic changes, Bloom and Williamson (1998), using standard and even rudimentary econometric techniques, found that approximately one-third of East Asia’s increase in per capita income was due one way or another to the impact of the demographic dividend. In a corpus of work which very much parallels that of the above mentioned Harvard economists, Andrew Mason (Mason, 2001a), using growth accounting rather than econometric methods, estimated that the 'demographic dividend' which Bloom and Williamson had identified accounted for about a quarter of the Asian Tiger’s economic growth during their 'growth spurt'. While not everyone has accepted these conclusions at face value (see eg Schultz, 2005), there is little doubt that an increase in the proportion of the population of working age, and an accompanying increase in 'prime age' workers, must have been a significant factor in the growth transition evidenced by these countries. There are sound and elementary (econ 101) type reasons for thinking that this ought to be the case, and there is growing empirical evidence that it was.

In the third place evidence of the importance of population change on growth performance has been coming from a new and unexpected source: the ageing developed economies. A growing number of developed countries now find themselves with a rapidly increasing proportion of their population over retirement age, lowest-low levels of fertility and continuing and unprecedently (for mature economies) low levels of economic growth sustained over a considerable period of time. The combination in time of these two processes ( falling fertility and increasing life expectancy) with the 'low growth impact' has meant that economists have started to 'prick up their ears' on the 'population topic' as never before.

While it should be obvious, it is perhaps worth noting here that all the above-cited authors when they argue that age-structure variables do have predictive power and can 'explain' a significant portion of economic growth during the development transition, continually stress that the relationship between demographic variables and economic development is not a deterministic one. Population matters, but policy matters to, and whether or not governments put in place an appropriate package of policies, or whether or not a country enters an adequate 'mindset' makes a big difference in deciding whether the opportunity offered by the demographic dividend is put to productive use or simply frittered. The difference between the experiences of some Latin American countries and some Asian ones is clearly salutory here.

It is also perhaps worth pointing out at this stage that a further possibility does, in fact, exist: that movements in the key policy variables like democracy, openness, defence of private property rights etc etc, may turn out to be endogenous to the transition process which the changing age structure sets in motion (in particular here see Chapter Two on Malmberg's idea of four phases in the age structure transition). This possibility will be discussed more fully later in this book.

The Coale Hoover Hypothesis

The recent blossoming of interest in the demographic and age-structure related components of growth in fact dates-back to 1958 and to the publication of what was at the time a highly influential book by Ansley J. Coale and Edgar Hoover: Population Growth and Economic Development in Low-Income Countries.

In their book Coale and Hoover advance what has subsequently come to be known as the Coale-Hoover hypothesis. This hypothesis is based on a simple but powerful intuition: rapid population growth arising from falling infant and child mortality swells the ranks of dependent young, and this single demographic event, in and of itself, increases the consumption share in national income at the expense of the saving one. Despite all the comings and goings around their hypothesis over the years, this single insight has survived more or less intact, and will surely form one of the core building blocks of all subsequent development theory.

By analysing simulation results derived from an econometric model which had been calibrated using Indian data, Coale and Hoover concluded that India's development would be substantially enhanced if there were lower rates of population growth. Their analysis here rested on two premises. Firstly, that in post initial-mortality-decline 'child-heavy' societies household and aggregate saving is reduced by the generalised presence of large families. And secondly, that the existence of such high ratios of dependent children skews aggregate investment away from the more self-evidently economically productive activities, since there is a continuous pressure for resources to be transferred towards so-called 'unproductive' population-sensitive social expenditures (like health and education).

The key novelty in the Coale Hoover model was the linking of this 'crowding out' process to the age composition of a high-fertility population, and not simply to its size, density, or growth, per se.

A pioneering work in many ways, their book:

1) identified several possible theoretical linkages between population growth and economic growth which were in perfectly compatible with the general of the time about the economic-growth process (which tended, for example, to place considerable emphasis on physical capital formation);

2) formalized these linkages into a generalised model that was parameterized and simulated to generate forecasts of alternative fertility scenarios over the intermediate-run;

3) offered an enlightening case study of a major developing country country whose short to mid term economic prospects prospects were considered to be far from good. The Coale-Hoover framework was transparent and easy to understand, the assumptions were explicit and qualified, and the findings were clearly expounded and accessible to a wide readership.

The model identified, and the simulations quantified, three adverse impacts of population growth:

1) A capital-shallowing impact: a rapidly growing working age population produces a reduction in the ratio of capital to labour since population growth per se does not increase saving rates, and a rapid increase in the proportion of young people tends to lower the aggregate savings ratio;

2) An age-dependency impact: an increase in youth-dependency raises the requirements for household consumption at the expense of saving, thus diminishing the savings rate;

3) An investment diversion impact: a shift of (mainly government) spending into areas such as health and education at the expense of (assumed-to-be) more productive, growth-oriented investments.

In particular, as has been suggested, the hypothesis attracted a good deal of attention from growth economists due to its emphasis on the role of physical capital (as distinct from the earlier Malthusian focus which was on land as a scarce resource). At the time physical capital accumulation was considered by many to be 'the' key to economic development.

Hoover and Coale's work, which had a substantial impact on U.S. population policy and thinking at the time of publication, did not however go unchallenged, even at the time of publication, and its influence was subsequently to wane rapidly as empirical research failed to uncover the strong and consistent impacts of population movements on saving in the developing and less developed world which the Coale Hoover hypothesis was interpreted as anticipating.

Initial research results did however seem be favourable. Nathaniel Leff, for example, in an early study based on a sample of 74 countries, found the log of gross savings rates to be inversely related to the proportion of the population either under 15 or over 64 (Leff, 1969), a finding which appeared to place the youth dependency hypothesis on a solid empirical footing. Subsequent research, however, (Goldberger,1973, Ram, 1982) failed to find confirmation of the dependency hypothesis and some researchers even cast doubt on the validity of econometric procedures employed in the initial Leff study.

Subsequent developments in growth theory also seemed to undermine the foundations of the dependency hypothesis. The principal rival to the hypothesis, Tobin’s life-cycle model (Tobin, 1967), took it as axiomatic that savings rates should increase as population growth did. The reason Tobin thought this is not difficult to comprehend, at least in the original version of his model which worked with a typology of only two population groups (workers and a retired population): faster population growth tilts the age distribution toward young, saving, households and away from older, dissaving ones. This way of looking at the process, which contains the germ of a very interesting idea in the 'tilt effect' it proposes, was, however, seriously misleading as it rested on a considerable over-simplification of the life cycle process.

The representative agent version of Robert Solow’s neoclassical growth model also lead thinking in a similar direction, with faster population growth being thought to result in higher savings rates, savings rates which go up as interest rates raise in response to heightened investment demand (Cass 1965, Phelps 1968, and Solow 1956). However, neither class of model explicitly addressed the dynamics of the dividend which Hoover and Coale had considered to form part and parcel of the ongoing impact of the demographic transition.

In Tobin’s steady-state model the "age tilt" factor, whose presence is interesting and insightful in and of itself, arises as a result of an initial modelling decision which described a world where there were only two groups of partyicipants, active adults and retired dependents. Had the model also incorporated the idea of youth dependency a very different tilt-effect would in all likelihood have been produced.

In similar fashion, the standard neoclassical growth models assume exogenously fixed labor participation rates, and, by implication, assume no endogenously driven changes in the dependency ratio. Clearly this kind of assumption is - strictly speaking - 'necessary' if one's objective is the conceptualisation of "steady-state" behaviour, but this precisely begs the question as to whether, in the light of the impact of key "transitional dynamics" during a continuing process of demographic change, the postulation of even a hypothetical 'steady state' economy is a useful and valid procedure. Certainly as we shall see later it is hard to find many examples (by 'many' I mean here enough to count on the fingers of one hand) of the kind of constant 'mature trend' which Kaldor presented as one of the core stylised facts typifying modern economic growth (Kaldor, 1961).

In effect it could be argued that by working backwards from the steady state idea the whole neoclassical school of models gain in modelling simplicity precisely at the expense of sacrificing the rich and complex population dynamics which were implicit in Coale and Hoover’s early theorising and which were to become so blatantly empirically evident in the course of East-Asian demographic transition.

That subsequent empirical findings were to go against Coale and Hoover was not perhaps entirely surprising in the context of the times. At the household level, the saving impacts they attempt to describe are fundamentally based on a 'life-cycle' conceptualization of behaviour, and such a conceptualisation requires a substantial 'forward looking' planning horizon. The behavioural transition which is required for this to occur also involves a considerable evolution in institutional structure (developed capital markets, reliable pension options etc) in order to make the implementation of such lifetime plans feasible. At the time Coale and Hoover were writing these conditions simply did not exist in the vast majority of the then third world countries. For many families living in an agricultural context spending on children represents an investment in a cheap ongoing stream of willing or unwilling labour (Kramer and Boone etc Edward, and Kaplans comparisons with different levels of skill formation in foraging societies) as well as a form of saving for an uncertain future (e.g., parents may expect transfers from their children in old age - think institutional structures, Lee etc : Edward) and, children, as many studies reveal, can be viewed as a productive asset both in the household and on the farm (Doepke 2004, Doepke and Zilibotti 2005, Kramer and Boone, 2002).

The research climate at the end of the 80s, and the evaluation of Coale and Hoover which this produced is perhaps well summed-up in a single observation from Angus Deaton (1992): "Although there are some studies that find ... demographic effects, the results are typically not robust, and there is no consensus on the direction of the effect on saving."

New-Wave Age Structure Theory

As has been said above, the revival of interest in the Coale-Hoover model which has taken place in recent years has been considerable , and this despite its known limitations (see eg Williamson 2001). In fact it would be no exaggeration to say that a revised form of their dependency hypothesis is now enjoying something of a renaissance. The original Coale-Hoover insight has been developed into a class of explicit economic models that, suitably calibrated, account tolerably well for the cross-country savings variations which are to be found in macro time-series for many developing countries. Almost all the recent analyses of developing countries macro data which have tested explicitly for age structure impacts find the presence of Coale-Hoover effects to a greater or lesser degree (Collins 1991; Harrigan 1996; Higgins 1994, 1998a; Kang 1994; Kelley and Schmidt 1995, 1996; Lee, Mason, and Miller 1997; Masson 1990; Taylor 1995; Taylor and Williamson 1994; Webb and Zia 1990; and Williamson 1993).

This renaissance of the Coale Hoover hypothesis has in large measure been possible due to the work of a fairly limited number of researchers. What might be reasonably be termed the 'new wave' era of age-structure research probably has its begining with an early paper by Mason and Fry (Mason and Fry, 1982). As subsequently elaborated-on by Mason (Mason 1987,Mason 1988), this work develops what Mason and Fry called a 'variable rate-of-growth effect' model which tries to establish a link between youth dependency ratios and national saving rates. The model relies principally on the insight that given the existence of positive labor productivity growth, younger cohorts enjoy the prospect of continually higher permanent incomes and as a consequence higher consumption than their parents did. If at the same time consumption is shifted from child-rearing to the later, non-childrearing stages of the life cycle, aggregate savings rise with a momentum that depends directly on the growth rate of national income. In their work the dependency and lifecyle perspectives are unified through a full blooded incorporation of the effect of changes in the youth dependency ratio on the timing of life-cycle consumption.

A decline in the youth dependency ratio, for example, should cause consumption to be shifted from the childrearing years to later, non-childrearing, stages of the lifecycle. As a result, following the Mason and Fry model, the saving rate depends on the product of the youth dependency ratio and the growth rate of national income ( ie they explicitly incorporate the Tobin 'growth-tilt 'effect), as well as on the dependency ratio itself (the 'composition effect').

This result had an important qualitative implication for the "classic" dependency model: the demographic "center of gravity" for investment demand is located earlier in the age distribution than the centre of gravity for savings supply. In particular, investment demand should be more closely related to the share of young people in a population (through its connection with labor-force growth), while savings supply should be more closely related to share of mature adults (through its connection with retirement preparations). The divergence between these centers of gravity also has the interesting implication that the effects of demographic change on savings and net capital flows will depend on the economy's degree of openness to capital flows (Higgins, 1994, 1998).

Simply put, in an open economy, a population with a heavily child-centred age distribution should exhibit a tendency towards current account deficits: savings are low due to the high youth dependency burden. Later as increasing numbers of young people enter the labour market investment rises in response to higher labour-force growth. Then as the age distribution shifts steadily upwards, the savings supply should increase pushing the current account into surplus (Cutler et al, 1990). Of course all such schematic ideas need to be subject to important qualifications. As Bo Malmberg argues, child dominated societies also tend to be heavily dependent on primary extractive industries, and hence on the price of raw materials, like say energy, or precious metals (Malmberg, 2000). Thus not all 'child rich' societies will experience current account deficits ( a fact which Brad Setser, for example, is only too happy to draw to our attention). Countries like Saudi Arabia, and other oil-rich 'young' states, may demonstrate a capacity for significant and sustained long term current account surpluses. Interestingly, in a global-macro context, these surpluses may be related to high demographic-dividend-driven growth in other parts of the global economy.

It is also important to note that the negative saving impacts found to be associated with increases in the share of the elderly population need not indicate that they themselves are actually drawing down their stocks of assets. The burden of supporting the elderly (either directly or through transfer payments) might lead to lower saving by younger households. Alternatively, prime-age households with elderly parents might save less in anticipation of bequest receipts (Weil, 1994). The age coefficients here are not then simple behavioural parameters which describe the actions of agents belonging to different age groups, but instead attempt to capture the relationship between age distribution and the behaviour of economic agents of all ages.

The Demographic Dividend

Recent use of the expression demographic dividend to describe the positive feedback mechanism associated with the fertility decline component of the demographic transition dates back, as we have seen, to the East Asian growth study of David Bloom and Jeffrey Williamson (Bloom and Williamson 1998). According to their argument as advanced in that study the "demographic dividend" leads to opportunities for growth of output per capita for two principal reasons.

In the first place there is an impact on total GDP due to a "growth accounting effect". A rising share of the total population in the working-age group increases the ratio of 'producers' to 'consumers'. Obviously this situation contributes directly and positively to a growth in output per capita.

Secondly, they conjecture that age-distributions might also be associated with what they call 'behavioral effects' (a growing proportion of prime age workers enhancing overall productivity, for example, and more people in the prime age group means a potentially higher personal savings rate and hence, potentially, lower interest rates) and these behavioural effects in their turn have a favourable influence on the growth of output per capita.

Formally the the accounting effect and the behavioral effect can be broken down in a fairly straightforward fashion between growth in the proportion of the population of working age and growth in the level of productivity per worker, although the former does not always and everywhere imply the latter since in part this depends on the educational and skill level of the new labour market entrants.

Simply stated, then, the demographic dividend occurs when a fall in the birth rate following an initial mortality decline produces changes in the age distribution of a society, and when these changes have the result that fewer investments are required to meet the needs of the youngest age groups. In this way resources are released which may be used for investment in economic development and for improved family welfare. That is, a falling birth rate makes for a smaller population at the young, dependent, ages and implies the presence of a relatively larger share of those in the adult age groups who comprise the productive labor force. It improves the ratio of productive workers to child dependents in the population. This situation generally makes for faster economic growth and fewer burdens on families.

In addition, the decline in fertility which accompanies the transition is also normally accompanied by an increase in female education levels and, concommitently, with an increasing level of female participation in the labour force (Lutz and Scherbov, 2004).

This demographic dividend, however, is not available forever. It operates through a limited time window. As the demographic transition continues its course, the age distribution changes again, and the large adult cohorts move inexorably into the older, less-productive age groups being followed in turn into the more productive ages by the smaller cohorts that were born later in the fertility decline. When this occurs, the dependency ratio rises again, but this time it is a question of growing care and support needs for the elderly, rather than the demands of child rearing and the economic consequences that these demands have.

It should also be stressed that the dividend itself is in no way automatic. While demographic pressures are normally eased when fertility initially falls, some countries will take better advantage of this easing than others. Some countries will act to capitalize on resources released and will use them effectively, while others will not. As, in the course of time, the favourable window once more closes, those that who have not found the way to take ample advantage of the demographic dividend may well face renewed resource pressures at a time when their ability to respond is weaker than ever. The current situation of the Eastern and Central European economies may well furnish one example of a flawed dividend situation, for while many of these economies are now, thanks to their recent incorporation in the European Union, belatedly closing the living standards gap which had opened up, they are also ageing very rapidly (and of course not all these countries by any means are in the EU: the Russian Federation, Belorussia, Ukraine, Moldava, Georgia etc remain outside, and will almost certainly grow old before they grow rich, if that is, they many to avoid that worst of all eventualities - population meltdown - on the way). Other examples of countries with below replacement populations which have yet to achieve economic development would be Cuba, North Korea and Iran. It may well be that in all these cases the earlier failure to capitalise will truly mean that many of these countries will grow but never grow rich, and the fact they missed the dividend boat may well mean they pay a high price for this in the form of a difficult and uncertain future.

The demographic dividend itself works through the combined operation of several interconnected mechanisms.

Labor Supply

As the demographic transition follows its course the generations of children born during the high fertility years enter adult life and become workers. Women who are now having fewer children than before are released from childrearing responsibilities and are able to take jobs outside of the home; also, as the transition moves forward, and years of compulsory education increase, younger women tend to become better educated than those to be found in the older cohorts, and are thus more productive once inside the labor force.

Savings

Mature working-age adults tend to earn more and thus are potentially bigger savers than are new labour market entrants, or those setting up an independent home for the first time. Thus the larger generations who work their way through the labour force as the age pyramid changes favour greater personal and national saving. This ability to save becomes even greater as the 'thick cohorts' move into their 40s, especially as in the first instance the generation-span is smaller, and their own children rapidly become wage-earners themselves and hence soon require less support. Thus personal savings continue to grow and are able to serve as a source of investment funds. Countries steadily move from being heavily dependent on external finance, to a position of relative financial self-sufficiency (the contemporary Chinese example, of course, immediately springs to mind here).

Human Capital

Having fewer children normally enhances the health of both mother and child. Female participation in the labor force, in turn, enhances the social status and personal and financial independence of women. Also fewer children normally means fewer and better educated ones. More investment is allocated to each individual child.

Theory and Modeling

The beginings of a more general awareness among economists of the possibility that such demographically related processes might have more importance for understanding economic growth than had been traditionally attributed to them coincided with the emergence in the 1990s of the empirical 'convergence' growth research. Pioneered by Robert Barro, this work postulates the existence of either a universal, or a country-specific, long-run steady state growth-rate for economies. In the former case economists tended to talk of absolute long run convergence, while in the latter they tended to talk about 'conditional convergence'. Based on this strong initial assumption researchers proceeded in an attempt to identify the factors (economic, political, social, institutional, geographic etc) that determine both the long-run steady state growth rate for individual countries, and, in the shorter-to-intermediate-run, the transition dynamics of each country towards this longer-run state. Conveniently the models lend themselves fairly readily to certain types of demographic testing due to in part to this very differentiation between short- and long-run impacts.

Three approaches have tended to characterise the now extensive literature on economic-demographic modeling: simple correlations, production functions, and convergence patterns (Kelley and Schmidt, 2005).

Simple-correlation studies normally hypothesize that per capita output growth is influenced by a variety of demographic dimensions in the following fashion: Y/Ngr = f(D), where Y is national income, Ngr is the rate of growth in labour supply, and D is normally some kind of proxy for demographic components, like Total Fertility Rates.

Production-function studies are based on estimating variants of a model which takes some variant of the form

Y = g (K, L, H, R, T), where K is capital, L is labour, H is human capital, T is technology etc

Convergence-pattern studies, are rooted in neoclassical growth theory, and explore the relationships between economic growth and the level of economic development. They focus on the pace at which countries move from their current level of labour productivity to their long-run, or steady-state equilibrium level of labour productivity.

One revealing feature of the convergence-pattern models can be clearly illustrated by taking a closer look at some of the variables which have been omitted by those working in the Barro tradition. In the majority of papers authors emphasize variables that determine longrun, or 'potential' labour productivity, and downplay variables that might be instrumental in bringing about the 'adjustment', or transition, to long-run equilibrium. An example of one such omitted variable would be the investment share of GDP.

Indeed investment might be considered to be the first variable one would think to include in a model which focuses on labour productivity since Levine and Renelt (1992) in an influential study surveyed a wide variety of empirical growth research in an attempt to identify a common set of influential variables and found, somewhat unsurprisingly, that investment rates constituted the single most robust variable. Rather than implying that the investment rate is a viable variable in-and-of itself to use in determining long-run capital-to-output ratios and through these long run trend growth potential, the significance of this finding is most probably that it suggests an incomplete set of variables are being tested. If the convergence hypothesis were correct and the list of Barro variables were complete, the investment coefficient should in theory be insignificant (see Bloom, Canning, and Malaney 2001, Higgins and Williamson 1994, and Kelley and Schmidt 1994).

One way of looking at this issue is to start from the proposition that the gap between current and long-run labor force productivity largely dictates the return to investment. According to standard theory investment should flow to those countries which exhibit the highest returns, so, rather than investment accounting for growth per se, it could be argued that the it is the 'structural' features of a country (among which one would evidently incorporate demographic structure) that either impede or facilitate investment, and thus that ultimately determine growth.

Attempts to incorporate demography into convergence models have been few, and in general relatively ad hoc. The demographic variables that normally qualify for inclusion are those that affect long run labour productivity, and those that condition the transition to it.

Barro himself seems to have been aware of the problem of neglecting demographic dynamics and did subsequently attempt to incorporate some element of demography into his use of the standard convergence models (Barro 1991). In a first pass at the problem he and his co-workers concluded that high fertility, strong population growth, and high mortality all exert negative impacts on per capita output growth. In 1994 Kelley and Schmidt extended the list of variables to include population density and size, which in fact revealed positive impacts, although the net assessment for the complete set of demographic variables tested was negative. The density finding is interesting, however, since it raises the question as to whether density itself serves as a proxy for some other 'hidden' variable which is not itself being tested for. This interest is only heightened by the recent finding of Wolfgang Lutz and his co-workers that human fertility declines as population density rises (Lutz et al 2005), and by applying to this the thought that both of these may be a reflection of changing female labour force availability and education levels.

As indicated above, Barro focuses on one single demographic variable, the Total Fertility Rate (Barro, 1997) . This variable captures both the adverse capital-shallowing impact of more rapid population growth, and the resource costs of raising children versus producing other goods and services. By its very nature the TFR exerts its impacts mainly on long-run population trends and on potential long run labour force growth, a focus which tends to understate the short- and mid-term transition dynamics (which may of course last for decades) en route to to the postulated equilibrium state. TFR is, it should always be remembered, a statistical construct that attempts to what population dynamics 'would be' if the current age-specific fertility rates were maintained over a long period of time. Measurement and interpretation issues abound here (Bongaarts and Feeney, 1998, Sobotka et al, 2005).

Rooted in neoclassical growth theory, the framework Barro uses explores the relationships between economic growth and the level of economic development of a country. It focuses on the pace at which countries move from their current output level to what is postulated as their long-run, or potential, or steady-state equilibrium level of output.

So the model from the outset contains a convergence assumption:

In Barro's model the rate of output growth per worker is proportional to the gap between the logs of the long-run, steady-state growth rate and the current one. The greater this gap, the greater are the gaps of physical capital, human capital, and/or technical efficiency from their potential levels. Large gaps allow for 'catching up' through (physical and human) capital accumulation and technology transfer across, and within, countries.

The rate of convergence is assumed to be independent of time and place. By contrast, potential steady-state output per worker is specific to country and time. This potential is, of course, unobservable and is simply a modelling heuristic. Its log is modeled as a linear function of a vector of country- and time-specific characteristics which normally take a form similar to:

ln(Y/Lit)* = a + b Zit.

The actual specification of the determinants of long-run labor productivity (i.e., the selection of of the variables which are included in Z ) varies considerably from one study to another, but the basic framework used is the same across a very large number of empirical studies:

What types of Z variables should be included as determinants of long-run output per worker, we may ask? Recognizing that a long-run, steady-state production function in principle lies behind the postulated steady state growth rate, factors which influence long-run physical and human capital stocks, technology, and natural resource stocks have been considered to be important. Also included among are factors like market structure, access to ports, climate, policies toward trade, education, health, etc.

However, not all countries finance investment with equal ease. Thus, a second category of variables has been added to the Z vector in order to 'condition' the convergence rate. These variables have included country- and time-specific factors which may be thought to either enhance or deter international capital flows, domestic saving, domestic investment, and/or migration. Included among these are, for example, restrictive licensing, the risk of expropriation, political conditions, the rule of law, migration regulations, and so forth.

In a number of works Barro has outlined a list of core economic Z variables (Barro, 1991, 1997). Growth in output per capita is held to be positively related to:

1/ A lower initial level of productivity. Convergence is posited as being more rapid the greater this difference and the the higher initial levels of schooling.

2/ Higher male secondary and tertiary schooling attainment, which facilitates the absorption of new technologies.

3/ Higher life expectancy, which is normally considered to be a proxy for better health and better quality of human capital in general.

4/. Movements in the terms of trade.

5/ Inflation levels, with lower inflation being thought to lead to better forecasting decisions based on more predictable price expectations.

6/ A lower government consumption share - net of education and defense spending - since this in theory releases resources for more productive private investment.

(7) The strength of democratic institutions, since these promote market activity by loosening 'rentier-type' controls, by facilitating more transparency, and generally weakening the hold of corruption. However this demographic component is normally conditioned according to the economic development of the country, since stronger democracies in weakly economically developed countries can dampen growth by exerting pressure for an increasingly active government role in redistributing income, and for higher spending on basic welfare services regardless of the capacity of the economy to support them (Viz, Bolivia at the present time).

7/ A stronger rule of law which stimulates investment by promoting the honouring of contracts, security of property rights, the presence of intellectual property rights etc, etc.

The Harvard School

As has already been intimated, the end of the 1990s saw a veritable boom in econometric testing of the 'demographic hypothesis' which built on the early example of Coale and Hoover. In the forefront of this boom were a number of prominent Harvard economists (among these David Bloom, David Canning, Jeffrey Sachs, and Jeffrey Williamson).

Building on the original Barro setup (albeit with a different choice of core variables), the Harvard researchers focused on the population impacts of the age-structure changes which are normally associated with the demographic transition. Their core model attempts to capture these impacts compactly by linking just two variables: population growth and working-age growth. Such a specification can be readily achieved by “translating” a traditional neoclassical growth model, which is formulated in output per worker terms, into an equivalent model formulated in output-per-capita growth terms. This translation has the virtue that it provides a straightforward way of highlighting some shorter-period “population impacts” within the usual long-run neoclassical framework. Numerous empirical papers by the Harvard economists have found significant demographic impacts, especially in the East Asian case. (more references)

Much of the Harvard work takes as its starting point the alternative methodology for exposing dynamic demographic relationships which was advanced in the original Bloom and Williamson paper (Bloom and Williamson, 1998, B&W hereafter).

In the Bloom and Williamson version of the demographic dividend thesis, as has been explained above, the accounting and behavioral effects of the age transition can be decomposed using fairly straightforward econometric techniques (ones which essentially involve the formulation of an identity function, the taking of natural logarithms of output per capita and the differentiation of the resulting expression with respect to time).

Starting with a definition of output per labour hour, B&W illustrate how the basic Barro model can be transformed into one which describes the growth process in per-capita terms. Now the impacts of working-hour growth and population growth just cancel each other out when they change at the same rate, something which would certainly occur in the hypothetical steady-state growth situation with a static age pyramid. This 'unworldy' assumption (unworldly since outside the rather unique case of the USA this phenomenon is virually unknown empirically) is normally - as has been indicated - imposed by fiat (it certainly cannot be induced empirically) in most standard convergence studies. B&W, on the other hand, take as their starting point the fact that in the developing world the demographic transition only really got started after WWI, and that the 1960s, 1970s, and 1980s continued to be decades of heavy demographic transition in most developing countries. As a result, the 'mutually cancelling' condition did not hold and differential growth rates should, in principal, be observed. This of course is presisely what in practice was found to be the case.

The basic B&W set-up does not model workforce-share and its relation to output growth. Quite the contrary, B&W replace labour force growth with a pure demographic proxy, the growth rate of the working-age population. They thus build their model as if the only determinant on hours worked were the age-distribution of the population, and hence the relative growth of the working-age versus the full population constitutes the sole impact of demography in their model. Following the B&W setup, sometimes the impact of demography will be positive, sometimes negative, and sometimes zero. The model thus highlights the reality that demographic impacts vary during the transition but is silent on the issue of possible demographic impacts on long-run labor productivity; i.e., demography does not affect the Barro convergence variables.

As a result, the BW model has a narrower interpretation than most other renderings in the more recent literature, which often admit both short- and long-run impacts of demographic change as a part of their theoretical structure. On the other hand, it has the desirable attribute of clarity in its interpretation. Thus the postulated relations between population growth and working age growth allow for a clear interpretation of the role of demography in the transition dynamics in that relatively rapid growth of the working-age population will speed the transition to long-run economic prosperity. At the same time two countries with the same convergence variables will ultimately arrive at the same level of labour productivity growth, irrespective of their demography. B&W acknowledge the possibility that the rates of growth of working age and total population might impact long run labour productivity, but at no point do they attempt to model this explicitly. To be continued.

References

Barro, Robert J. 1997. Determinants of Economic Growth: A Cross-Country Empirical Study. Cambridge, Mass.: MIT Press.

Barro, R.J. 1991. “Economic Growth in a Cross Section of Countries.” Quarterly Journal of Economics 106(2):407-444

Barro, Robert J. and Jong-Wha Lee. 1993. "Losers and Winners in Economic Growth." NBER Working Paper

Barro, Robert J, and Xavier Sala i Martin, 2003, Economic Growth (Second Edition), Cambridge, MIT Press

Birdsall, Nancy, 2003, "New Findings in Economics and Demography: Implications for Policies to Reduce Poverty" in Population Matters: Demographic Change, Economic Growth, and Poverty in the Developing World Edited by Nancy Birdsall, Allen C. Kelley, Steven W. Sinding, Oxford, Oxford University Press

Bloom D, Canning D, Moore M (2004). The effect of improvements in health and longevity on optimal retirement and saving. NBER Working Paper No. 10919. Cambridge, National Bureau of Economic Research.

Bloom D, Canning D, Graham B (2003). Longevity and life-cycle savings. Scandinavian

Journal of Economics, 105:319-338.

Bloom, D. E., D. Canning and Sevilla (2002). The Demographic Dividend: A New Perspective on the Economic Consequences of Population Change. Santa Monica, CA, RAND.

Bloom, D.D., Canning, D. & Malaney, P. (2001). Demographic change and economic growth in Asia, Population and Development Review, 26, supp., 257{290.

Bloom, D. E. and J. G. Williamson (1998). Demographic Transitions and Economic Miracles in Emerging Asia. World Bank Economic Review vol.12, No.3, pp. 419-56.

Bongaarts, J. and G. Feeney. 1998. “On the quantum and tempo of fertility,” Population and Development Review 24(2): 271–291.

Coale, Ansley J. and Hoover, Edgar M. 1958. Population Growth and Economic Development in Low-Income Countries. Princeton: Princeton University Press.

Collins, Susan. 1991. “Saving Behavior in Ten Developing Countries.” In B. Douglas

Bernheim and John B. Shoven, eds., National Saving and Economic Performance.

Chicago: University of Chicago Press.

Cutler, D. M., J. M. Poterba, L. Sheiner and L. Summers 1990. "An Aging Society: Opportunity or Challenge?" Brookings Papers on Economic Activity No. 1, pp.1-56.

Deaton, A. and C. H. Paxson (2000). Growth, demographic structure, and national saving in Taiwan. In Population and Economic Change in East Asia, A Supplement to Population and Development Review. R. Lee and C. Y. C. Chu, eds. New York, Population Council. vol. 26, pp.141-173.

Doepke, M. 2004, “Accounting for Fertility Decline During the Transition to Growth,” Journal of Economic Growth, September 2004, 9 (3), 347–83.

Doepke, M and F. Zilibotti 2005, "The Macroeconomics of Child-Labor Regulation.". American Economic Review, December 2005.

Fry, Maxwell, and Andrew Mason. 1982. “The Variable Rate-of-Growth Effect in the Life-Cycle Saving Model: Children, Capital Inflows, Interest, and Growth in a New Specification of the Life-Cycle Model Applied to Seven Asian Developing Countries.” Economic Inquiry 20(3, July):426–42.

Goldberger, Arthur. 1973. “Dependency Rates and Savings Rates: Further Comment.”

American Economic Review 63(1, March):232–33.

Harrigan, Frank. 1996. “Saving Transitions in Southeast Asia.” Asian Development

Bank, Manila. Processed.

Higgins, Matthew 1998. “Demography, National Savings, and International Capital Flows.” International Economic Review 39(2, May):343–69.

Higgins, Matthew. 1994. “The Demographic Determinants of Savings, Investment, and International Capital Flows.” Ph.D. diss., Harvard University, Cambridge, Mass.

Higgins M, and Williamson J (1997). Age structure dynamics in Asia and dependence on foreign capital. Population and Development Review, 23:261-293.

Higgins, Matthew, and Jeffrey G. Williamson. 1996. “Asian Demography and Foreign Capital Dependence.” NBER Working Paper 5560. National Bureau of Economic Research, Cambridge, Mass.

Jones, C. 2001(a). Was an Industrial Revolution Inevitable? Economic Growth Over the Very Long Run" Advances in Macroeconomics, August 2001, Vol. 1, No. 2, Article 1.

Jones, C. 2001(b). Introduction to Economic Growth (Second Edition), New York, W. W. Norton & Company

Kang, Kenneth. 1994. “Why Did Koreans Save So Little and Why Do They Now Save

So Much?” International Economic Journal 8(4, winter):99–111.

Kelley, A.C. 2003, "The Population Debate in Historical Perspective: Revisionism Revised" in Population Matters: Demographic Change, Economic Growth, and Poverty in the Developing World Edited by Nancy Birdsall, Allen C. Kelley, Steven W. Sinding, Oxford, Oxford University Press

Kelley, Allan C, and Robert M. Schmidt, 2005. Evolution of Recent Economic-Demographic Modeling: A Synthesis, Journal of Population Economics, Springer-Verlag GmbH, Volume 18, Number 2, Pages: 275 - 300

Kelley, Allen C., and Robert M. Schmidt. 1995. “Aggregate Population and Economic Growth Correlations: The Role of the Components of Demographic Change.” Demography 32(4):543–55.

Kelley AC, Schmidt RM (1996). Saving, dependency, and development. Journal of Population Economics, 9:365-386.

Kinugasa, T. 2004. Life Expectancy, Labor Force, and Saving, Ph.D. Dissertation. University of Hawaii at Manoa

Kramer, Karen and James L. Boone, 2002. “Why Do Intensive Agriculturalists Have Higher Fertility? A Household Labor Budget Approach.” Current Anthropology 43(3):511-517

Kremer, M. 1993. "Population Growth and Technological Change: One Million B.C. to 1990" (Quarterly Journal of Economics, Volume 108 issue 3, August 1993, 681-716

Krugman, Paul. 1994. “The Myth of Asia’s Miracle.” Foreign Affairs 73(6, November-

December):62–78.

Kuznets, Simon. 1960. "Population Change and Aggregate Output" in Demographic and Economic Change in Developed Countries, Princeton, NJ, Princeton UP.

Lee RD, Mason A, Miller T (2000). Life-cycle saving and the demographic transition: the case of Taiwan. Population and Development Review, 26(Supplement):194–219.

Lee, Ronald, Andrew Mason, et al. (2003). From transfers to individual responsibility: Implications for savings and capital accumulation in Taiwan and the United States. Scandinavian Journal of Economics vol.105, No.3, pp.339-357.

Lee, Ronald, Andrew Mason, and Timothy Miller. 1997. “Saving, Wealth, and the Demographic Transition in East Asia.” Paper presented at the conference on population

and the Asian economic miracle, East-West Center, Honolulu, Hawaii, 7–10 January.

Processed.

Leff NH (1969). Dependency rates and savings rates. American Economic Review, 59: 886-896.

Lutz, Wolfgang, Maria Rita Testa and Dustin Penn, 2005. Fertility Declines with Higher Population Density, paper presented to the IUSSP XXV International Population Conference, Tours, France

Lutz, W. and S. Scherbov, 2004. Probabilistic Population Projections for India with Explicit Consideration of the Education-Fertility Link, International Statistical Review (2004), 72, 1, 81-92, The Netherlands.

Malmberg, Bo & Lena Sommestad. (2000). Four Phases in the Demographic Transition, Implications for Economic and Social Development in Sweden, 1820-2000. Arbetsrapport/Institutet för Framtidsstudier; Working Paper 2000:6

Mankiw, N. G., D. Romer, and D. Weil, (1992), “A Contribution to the Empirics of Economic Growth,” Quarterly Journal of Economics, 107, 2, 407-37.

Mason A (1988). Saving, economic growth, and demographic change. Population and Development Review, 14:113-144.

Mason A (1987). National saving rates and population growth: a new model and new evidence. In: Johnson DG, Lee RD, eds. Population growth and economic development: issues and evidence. University of Wisconsin Press, Madison.

Mason, Andrew (1988). Saving, economic growth, and demographic change. Population and

Development Review vol.14, No. 1, pp. 113-44.

Mason, Andrew, Thomas Merrick, and R. Paul Shaw, eds., 1999. Population Economics, Demographic Transition, and Development: Research and Policy Implications, WBI Working Papers (Washington, DC: World Bank Institute).

Mason, Andrew (2001). Population Change and Economic Development in East Asia: Challenges Met, Opportunities Seized. Stanford, Stanford University Press.

Masson, Paul. 1990. “Long-Term Macroeconomic Effects of Aging Populations.” Finance and Development 27(2, June):6–9.

Myrdal, Gunnar. 1940. Population: A Problem for Democracy. The Godkin Lectures, 1938. Cambridge, Mass.: Harvard University Press.

Perlman, M. 1975, Some Economic Growth Problems and the Part Population Policy Plays. Quarterly Journal of Economics, 1975 Volume 89, No2, Pages: 247-56

Persson, Joakim, 2004. The Population Age Distribution, Human Capital, and Economic Growth: The U.S. states 1930-2000, Department of Economics and Statistics, Örebro University, Sweden, Working Paper

Ram, Rati. 1982. “Dependency Rates and Aggregate Savings: A New International Cross-Section Study.” American Economic Review 72(3, June):537–44.

Reddaway, William Brian. 1939. The Economics of a Declining Population. London: Allen and Unwin.

Schultz, T. Paul, 2005, Demographic Determinants of Savings: Estimating and Interpreting the Aggregate Association in Asia, IZA Discussion Paper No. 1479, January 2005

Sobotka, Tomáš, Wolfgang Lutz, and Dimiter Philipov. 2005. 'Missing Births': Decomposing the Declining Number of Births in Europe into Tempo, Quantum, and Age Structure Effects. European Demographic Research Papers 2 , Vienna Institute of Demography, Vienna

Spengler, Joseph J. 1938. France Faces Depopulation. Durham. Duke University Press.

Taylor, Alan. 1995. “Debt, Dependence, and the Demographic Transition: Latin America into the Next Century.” World Development 23(5, May):869–79.

Taylor, Alan, and Jeffrey G. Williamson. 1994. “Capital Flows to the New World as an Intergenerational Transfer.” Journal of Political Economy 102(2, April):348–69.

Tobin, James. 1967. “Life-Cycle Savings and Balanced Economic Growth.” In William Fellner, ed., Ten Essays in the Tradition of Irving Fischer. New York: Wiley Press

Tsai IJ, Chu CYC, Chung CF (2000). Demographic transition and household saving in Taiwan. Population and Development Review, 26(Supplement):174-193.

Webb, Steven, and Heidi Zia. 1990. “Lower Birth Rates = Higher Saving in LDCs.” Finance and Development 27(2, June):12–14.

Weil, David N. 1994. The Saving of the Elderly in Micro and Macro Data, Quarterly Journal of Economics, Vol. 109, No. 1, pp. 55-81

Williamson, J.G. (2001), Demographic shocks and global factor flows. In J.N. Little and R.K. Triest eds. Seismic Shifts: The Economic Impact of Demographic Change (Boston, Mass.: Federal Reserve Bank of Boston).

Williamson, Jeffrey G. 1993. “Human Capital Deepening, Inequality, and Demographic Events Along the Asia-Pacific Rim.” In Gavin Jones, Naohiro Ogawa, and Jeffrey G. Williamson, eds., Human Resources and Development along the Asia-Pacific Rim. New York: Oxford University Press.

Young A (1994). Lessons from the East Asian NIC’s: a Contrarian view. European Economic Review, 38:964-973. Young A (1995). The tyranny of numbers: confronting the statistical realities of the East Asian growth experience. Quarterly Journal of Economics, 110:641-680.

Growth Tigers and Demographic Dividends

2008-02-10T13:02:00.000-08:00

The East Asian Tigers

The East Asian development experience has often been central in placing the issue of age structure back on the economic agenda, and is certainly the example which is most frequently cited in support of the age transition demographic dividend hypothesis, since in many ways the south east Asia "tiger" countries provided the first modern example of relatively poor economies finding the path to relatively rapid economic development (although astute observers might have been struck by importance of the earlier Japanese example),

In the case of the 'tigers' the most striking feature is how, as the transition progressed, a successful export-oriented growth-strategy produced more than enough jobs to absorb what was a rapidly growing workforce. At the same time, a relatively stable macroeconomic environment – at least until the financial crisis of the late 90s – provided what seemed to be a fertile and attractive investment environment.

Ongoing work by a number of researchers has continued to find that a significant part of the impressive rise in Asian savings rates can be explained by the equally impressive decline in dependency burdens, and that some at least of the notable difference in savings rates between what was in the 1990s a sluggish South Asia and a booming East Asia can be attributed to differences in their relative dependency burdens. If this view is accurate then it also fuels the rather optimistic hypothesis that some of the savings gap between the two regions should reduce as youth dependency rates fall in South Asia over the next two to three decades.

The debate over the East Asian "miracle" really begins with the work of Alwyn Young, who, in a couple of extraordinarily provocative papers (Young, 1994, 1995), attempted to demonstrate that the rapid economic growth that was only too evident in the region was principally attributable to increases in factor inputs - notably labor, capital, and education - rather than to general improvements in total factor productivity. If Young was right, then the key to understanding the rise in income levels in East Asia lay in understanding the driving mechanisms behind such a growth in inputs.

All of the Asian Tiger economies enjoyed a surge in savings and investment during the high-growth period. The private savings rate in Taiwan, for example, rose from around 5% in the 1950s to well over 20% in the 1980s and 1990s. (It is perhaps worth noting that much of the debate surrounding the savings issue has focused on Taiwan since the data on household savings in Taiwan is fairly comprehensive).

As predicted by the life cycle hypothesis, savings rates in Taiwan were found to vary by age, being highest for households whose heads were in the 50-60 age range. Using this data Bloom and Williamson found changing age structure to be a plausible explanation for the increase in aggregate saving which took place. In support of their argument they cited a number of previous studies (both "historic" and more recent ones) that had also found a strong connection between these two variables (Fry and Mason, 1982; Higgins, 1998; Higgins and Williamson, 1997; Kelley and Schmidt, 1996; Leff, 1969; Mason, 1987, 1988).

In terms of prior work, Higgins and Williamson (1996, 1997) had estimated the largest demogrphic macro impacts. Higgins and Williamson attempted to estimate the effect of changes in population age distribution on changes in, rather than levels of, the savings rate as it deviated around the 1950–92 mean. Thus East Asia’s savings rate in 1990–92 was 8.4 percentage points above its 1950–92 average due to the transition to a much lower dependency burden. Similarly, East Asia’s savings rate in 1970–74 was 5.2 percentage points below its 1950–92 average due, in part, to the size of the dependency burden.

They found the value of the total demographic swing to be a sizeable 13.6%, a swing which would appear to account for almost the entire rise in the savings rate in East Asia over the 20 years in question. The figures for Southeast Asia are similar, but not quite so dramatic. Southeast Asia’s savings rate was 7.9% higher in 1990–92 than the 1950–92 average and the mean was 3.6% between 1970–74. The total demographic swing was 11.5%, a smaller figure than for East Asia, but still apparently sufficient to account for the entire post 1970 rise in the Southeast Asia savings rate. Since the demographic transition had been slower in South Asia, the more modest changes in the savings rate were, more or less, to be anticipated.

Higgins (1998) also carried out an econometric investigation of the relations between national age distributions and savings and investment rates for a sample of 100 countries, using both time-series and cross-section data. Once more the results point to substantial demographic effects, with increases in both the youth and old-age dependency ratios associated with lower saving rates.

Perhaps more important for the agenda it would set for subsequent work, the results pointed to differential demographic effects on savings supply and investment demand, and thus, to a role for demography in determining the residual between the two: net capital flows or the current account balance.

In particular, Higgins found that high youth dependency rates demonstrated a strong relation with current account deficits, with young nations being found to join the ranks of the capital exporters as they mature. The estimated demographic effect on the current account balance had exceeded four percent of GDP over the last three decades in many of the 100 countries studied.

This Hiigins paper (which was based on his earlier doctoral thesis: he was one of Jeffrey Williamson's students) is best interpreted as extending the youth-dependency thesis Coale and Hoover (1958) had advanced almost forty years earlier, moving-on from an exclusive focus on saving rates to an analysis which incorporates the knock-on impacts for investment and the current account balance.

His finding that maturing nations tend to graduate from reliance on capital imports also has important implications concerning the possible evolution of the global savings pool in the decades which lie ahead. In particular, currently developing nations should be expected to be increasingly able to finance their own investment needs — a prospect which stands in in stark contrast to the pessimistic prognoses of the early 1990s ( For example, Depak Lal, 1991)

Higgin's other major result offered further confirmation that population dynamics accounted for a substantial share of East Asia’s economic miracle - a component equivalent to between 1.4 and 1.9 percentage points of East Asia’s total annual growth in GDP per capita from 1965 to 1990 (or as much as one-third of observed economic growth during the period) . Using a modified definition of "miracle" (and assuming, as Higgins does, that hypothetical "steady-state" growth in East Asia was a rate of about 2% a year) as being everything in excess of steady state growth, or about 4.1% (subtracting the notional 2 from the acieved 6.1 % growth rate). Population dynamics could then account for almost half of this observed difference or miracle. Now accounting for one-third or one-half of the East Asian growth certainly does not explain everything, but it does suggest that population dynamics may have been the single most important determinant of this 'extra' growth.

Within Asia itself, the evidence also suggests that demographic divergence contributed significantly to economic divergence during this same period.

Deaton and Paxson (2000) however throw something of a bucket of cold water on all of this since, using household savings data for Taiwan, they find that changes in age structure account for only a modest increase in the overall savings rate, perhaps 4 percentage points. They argue that the rise in the aggregate savings rate has not been mainly due to changes in the age composition of the population but, rather, to a secular rise in the savings rates of all age groups.

Accepting this finding at face value, the question then arises as to why savings rates should have risen at each age. One possible - and intriguing - explanation, proposed by Lee, Mason, and Miller (2000) is that increased savings rates are due to rising life expectancy and an increasing need to fund retirement income. In support of this idea Tsai, Chu, and Chung (2000) show that the timing of the rise in household savings rates does indeed match the recorded increases in the life expectancy of the population.

With a fixed retirement age we would expect such a savings effect. However, Deaton and Paxson (2000) argue that in a flexible economy, without mandatory retirement, the main effect of a rise in longevity will be on the span of the working life, with no obvious prediction for the rate of saving. Bloom, Canning, and Moore (2004) formalize this argument to show that under reasonable assumptions the optimal response to an improvement in health and a rise in life expectancy is to increase the length of working life, though less than proportionately, with no need to raise saving rates at all (due to the gains from enjoying compound interest over a longer life span).

Schultz (2005) has also argued that the estimated magnitude of the dynamic aggregate relationship identified by Higgins and williamson may be appears smaller than reported, as well as suggesting that the HW effect may be sensitive to the choice of econometric methods used to describe it. referring directly to the Deaton and Paxson finding that savings behavior at the household level does not demonstrate sufficient life cycle variation, he asks whether there might not be alternative explanations for the observed trends in savings? One possibility explanation for the empirical regularities would be the substitution of savings for children, and this in fact could be explained within a household lifetime demand framework.

There may also be a demographic effect as a longer prospective life span can change life-cycle behavior, leading to a longer working life or higher savings for retirement. Recent empirical literature on economic growth does indeed provide compelling evidence that health status, longevity and the age distribution do indeed have strong connections to economic growth (Bloom, Canning, and Graham, 2003; Bloom, Canning, and Moore, 2004, Kinugasa, 2004, Kinugasa and Mason, 2005, Mason 2005).

While in theory a longer life span should be associated with a longer working life, in practice this may not be the case. Bloom, Canning, and Graham (2003) find that, even allowing for age structure effects, longer life expectancy is strongly associated with higher national savings rates across countries, which suggests that there is a savings effect.

While the optimal response with perfect markets may be for workers to have a longer working life as their health improves and they have longer life expectancies, mandatory or conventional retirement ages, coupled with the strong financial incentives to retire that are inherent in many social security systems, seem to result in early retirement and increased needs for saving for old age.

Also old age "dependency" is something of a misnomer. Lee (2000) shows that, in all pre-industrial societies for which he was able to assemble evidence, the flow of transfers is from the middle-aged and old to the young. In developed countries, on the other hand, both the young and the old benefit from government transfers, and the net pattern of transfers is towards the elderly. However, at the level in the United States and elsewhere, elderly households in fact make significant transfers to middle-aged households, undoing to some extent the effects of government policy. This seems to suggest that the dependency burden of the elderly is a function of the institutional welfare systems that are in place rather than an immutable state of affairs.

Beyond East Asia, a number of other studies appear to have leant more support to the demographic dividend idea. Bloom and Canning (2003a), for example, look at the case of Egypt. Between 1965 and 1990 Egypt’s working-age population grew at the rate of 2.61% per annum. This rate was about about 30 % faster than the growth rate in the dependent population and is closely comparable to the corresponding rates of population change among slow-growing South Asian countries. By contrast, the working-age population of the East Asian countries grew at approximately ten times the rate of the dependent population during the "miracle" years, as the "baby bulge" cohort entered the prime-age worker group. In East Asia this was also associeted with a drop in fertility beyond the replacement level. Using simple econometric regressions Bloom and Canning estimated that - not unexpectedly - during the early phases of Egypt’s demographic transition age changes contributed a modest, but not insignificant, 0.4 percentage points to Egypt’s economic growth rate during 1965–1990.

The Celtic Tiger

Ireland is a small, open, trade-dependent economy and is one of the fastest growing economies in the developed world. It constitutes around 1.8% of overall output in the Eurozone. Its relatively high level of openness is reflected in the degree of international mobility of labour and capital in the Irish economy as reflected by both strong migrationary flows and high levels of foreign direct investment. Its high level of susceptibility to movements in external trade is signalled by a high share of combined exports and imports of goods to Gross Domestic Product (GDP) which was just under 150% in 2004 and the fact that its GNP is significantly lower than its GDP.

In recent decades the Irish economy has been transformed from being an agricultural and traditional-manufacturing based one, to one which is increasingly focused on the hi-tech and internationally traded services sectors. In 2004, the services sector accounted for 66% of total employment, industry for 28% and agriculture for just 6%. Over the last decade, unprecedented economic growth has seen the level of Irish real GDP almost double in size.

There have been many reasons advanced for Ireland's success, and these include EU membership and access to the the European Single Market; low corporation tax rate and a large multinational presence, a high proportion of the population of working age; increased participation in the labour market especially by females; a reversal of the trend of emigration toward immigration; sustained investment in education and training. Of these, the last four are clearly all factors which relate to Ireland's recent and somewhat belated demographic transition.

Throughout the twentieth century, Ireland was widely recognized as a demographic outlier in Western Europe. The Irish demographic regime was characterized by high rates of out migration, especially female migration, late marriage, high rates of permanent celibacy, low rates of cohabitation, low rates of non-marital fertility, and very high rates of marital fertility. Although Ireland did experience some decline in fertility in the first half

of the twentieth century, the decline was considerably less than those that took place elsewhere in Europe. In 1950, in spite of continuing late marriage and substantial permanent celibacy, Irish total fertility was still 3.3 (a level which was among the highest in Europe). Throughout the 1970s, Irish total fertility rates were still above three births per woman level, however, during the 1980s Irish fertility began to change dramatically. Coleman (1992) and Murphy-Lawless (1997) have chronicled this rapid change which meant that by the end of the decade fertility was approaching replacement level which is where it has stayed more or less to the present time.

Between 1976 and 1994 Ireland managed to increase its real income per adult from 50 percent of the U.S. level to some 60 percent. From 1994 onwards things changed surprisingly quickly, and its relative performance literally took off. It rose by 18 percentage points during the past six years to 2000 to reach 78 percent of the U.S. level.

In the case of Ireland, the decomposition of real GDP per working-age adult into its productivity and

employment-rate components brings out a startling fact. The growth in Irish productivity has been very

rapid, not just in the past few years, but over the entire 1976-2000 period, averaging 3.3 percent annually.

Productivity growth rates of 3 percent or more sustained over such an extended period have been a rare

occurrence in the postwar period, and particularly during the last quarter century. In fact, since 1975,

only Korea (among OECD member countries) has experienced faster productivity growth than Ireland.

However, if the Irish boom of 1994-2000 was not due to an acceleration of productivity (more output per worker),

it must logically be attributed to the other source of growth in GDP per adult, namely an exceptionally

strong increase in the employment rate (a larger fraction of adults at work). Like many other EU member countries Ireland suffered a major employment slowdown during the years between the mid-1970s and the mid-1990s, a feature which largely overshadowed its positive productivity performance. Since 1994, however, the burst of employment in Ireland has not only erased previous job losses, but it has pushed the country’s employment rate to levels which are now significantly above the European average. (Fortin, 2002).

The Irish employment boom has had little parallel in postwar Europe. Some European countries - Austria,

Denmark, the Netherlands, Norway, Portugal, and the United Kingdom - did see their unemployment rate fall appreciably during the decade.of the 90s, but none of them dropped from such a high initial level (16 percent in 1993 to less than 5 percent today) or so rapidly. Second, the dramatic increase in employment was able to draw on a very large pool of women who had never previously been in the labour force. The number of Irish women in the labour force has increased by 65 percent since 1993. Third, the rate of job creation has absorbed a very large flow of immigrants who were attracted (or attracted back) to Ireland by the boom.

As indicated above, since 1976 the growth rate of Irish productivity (output per employed worker) has been strong, registering an averaged rate of 3.3 percent per year. One important important factor underlying this increase has been the continuous shift of economic activity and employment from the relatively low productivity primary sector to the more productive parts of the secondary and tertiary sectors. A measure of the change is to be seen in the fact that Ireland’s primary sector was still employing 40 percent of workforce in 1960; a percentage had fallen to a mere 9 percent by the end of the century.

Another factor worthy of note in the Irish case has been the fact that, from the 1960s onward, Ireland has witnessed a rapid and radical change in education policy and priorities. in particular succesive governments have sought to facilitate free secondary and low-cost higher education. Interacting with a late baby boom produced as a by product of the fertility transition, this policy move has made available a plentiful supply of well-educated young workers and indeed Ireland now has one of the highest educational participation rates in the world - 81% of Irish students complete secondary-level education and approximately 50% go on to participate in higher education.

From 1960 to 1990, the growth rate of income per capita in Ireland was approximately 3.5 percent per annum. In the 1990s, the growth rate jumped to 5.8 percent, which is well in excess of any other European economy. This boost in the growth rate coincides closely with the falling dependency rate in Ireland. Thus, the raw data are consistent with the view that demographic change contributed to Ireland’s economic surge in the 1990s. Bloom and Canning (2003b) examine this argument more closely and argue that the economic boom that occurred in Ireland in the 1990s is well predicted by estimates derived from a standard age-structure econometric model. As part of their analysis, they also show how the growth in the working age to total population ratio was matched by the increase in labor supply per capita.Economic growth in Ireland was also fueled by two additional demography-based factors that increased labor supply per capita: (i) between 1980–2000 there was a substantial increase in female labor force participation rates, particularly in the 25-40 year old age group, and (ii) the decline in youth cohort sizes and rapid economic growth of the 1990s led to a reversal of the traditional outward migration flow, resulting in net in-migration of workers, made up partly of return migrants but also for the first time of substantial numbers of foreign migrants.

It is important to note that Ireland, like the “miracle” economies of East Asia, had in place economic and social policies that favored its taking advantage of the demographic shifts it experienced. Two key policies appear to have been at work in Ireland. First, in the late 1950s, there was recognition that the “closed economy” model of development had failed in Ireland. This led to new policies with an emphasis on encouraging direct foreign investment in Ireland and promoting exports. Second, from the mid-1960s, free secondary education was introduced, leading to a large increase in school enrollments and subsequent expansions in higher education. The resultant high levels of education, combined with export-oriented economic policies, seem to be powerful factors in ensuring that the benefits of the demographic transition are realized.

Turkey: Europe's New Tiger?

The Turkish economy has, in the past, notoriusly suffered from very high levels of macroeconomic instability. However, following an extraordinarily dramatic economic crisis in the years 2000-01, and a subsequent series of structural and financial reforms Turkey has suddenly converted itself into Europe's fastest growing economy, and at the time of writing (January 2006) the outlook for future economic stability and medium-term economic growth look extremely good indeed.

In the years between 1994 and 2001 Turkish economic growth averaged a mere 2.8% and the economy experienced a number of severe crises. However since early 2001, in an uninterrupted output expansion which has now lasted 15 quarters output per worker in Turkey's manufacturing sector has increased by 37.8%, lowering unit labour costs by 38.8%, in the last four years and dramatically improving Turkish competitiveness in international markets. The rate of total factor productivity growth in Turkey surged from 0.5% per annum in the 1990s to the impressive rate of 4.8% per annum achieved over the last four years: This productivity spurt provides the bulk of the explanation for the surprising newfound resilience of the Turkish economy and is able to account for no less than 55% of the rise in real GDP that has taken place. And this growth in GDP has been impressive, with real GDP growing at an annual rate of 7.4% over each the last fifteen quarters. In this sense it can safely be asserted that Turkey's productivity revival reflects a permanent shift toward a higher growth plateau and forms crucial part of the feedback-loop between economic stability, the declining cost of capital, and the increase in investment spending

At the same time it is important to remember that Turkey is still a relatively poor country and that a comparatively large share of its population is still employed in agriculture.Turkey’s GDP per capita income is low, but its economic catchup potential is considerable. Measured at market prices, Turkish GDP per capita was $3,400 in 2003, a level comparable to that of Bulgaria and Romania, and far below say even that of the Czech Republic and Hungary, who each roughly have a GDP per capita of around $8,300. However, a more accurate measure of per capita wealth is normally thought to be GDP per capita measured on a purchasing power parity (PPP) basis. Measured in this way Turkey’s per capita income amounts to around $6,700, again, broadly comparable to the current EU accession candidates (Bulgaria, Romania), and a level which is only around 20-25% of that of the largest EU economy, Germany. A low per capita income is a double edged issue, since it means a low standard of living and it suggests there is substantial room for “economic catch-up”, which implies considerable room to improve productivity through the virtuous circle of technological innovation and investment.

Alongside all this growth, Turkey's consumer price inflation rate has also declined sharply from an average of 77.5% in the 1990s to an 7.7% annual rate of in December 2005 ( single digit inflation for the first time in three decades). Estimates of inflation for 2006 and 2007 are also farly benign (in the 4% range for 2006 and the 3.5% range for 2007) and the Turkish central bank is thinking of setting inflation targets, 5% for 2006 and 4% for 2007 and 2008. The fact that inflation targeting is under consideration is indicative of the new institutional framework for monetary and fiscal policy which has been introduced. To this should be added extensive product, labour and financial markets reform, while the more selective use of infrastructure industry and agricultural support has added yet another dimension to a process which has effectively opened a window of opportunity which may well allow Turkey to genuinely escape from the boom-and-bust cycle of the past, and embark on a path of sustainable higher growth and stronger employment..

The ‘age of stability’

Overall, demographic developments in Turkey are now strongly conducive to higher economic growth. The last two decades have seen a continuous fertility decline and Turkey is now on the point of entering entered into the last - or 'second transition' - phase of its demographic transition. The latest nationwide Turkey Demographic and Health Survey (TDHS) reveals the current TFR as being close to reproduction level, although there is still a clear and significant west-east regional disparity, with the more modern Western part of Turkey now well inside the below-replacement camp.

Turkish population trends can be immediately seen by taking a quick glance at recent changes in some basic demographic indicators. The Turkish population increased by a factor of five over the last 70 years, with improvements in health services and living standards bringing about a steady decline in both child and adult mortality. The Crude Death Rate declined from around 30 per thousand in the 1940s to 7 per thousand at the start of the new millenium. (Yavuz, 2005). The second half of the 20th century also witnessed dramatic declines in fertility rates. In the early 1970s, the Total Fertility Rate was around the 5 children per woman mark, while current fertility as estimated in the last - 2003 - Turkey Demographic and Health Survey is thought to be just above replacement (2.2). As a result, the age structure of the population has been changing rapidly .

In addition, since the early 1950s, Turkey has also experienced substantial internal and external migration which has been associated with an extensive urbanization process. This urbanization is ongoing and continues to profoundly change the spatial distribution of the Turkish population which is now predominantly concentrated in urban settlements. Intensive migration between regions, in particular from the Kurdish east to the west and south, and from the interior to the coastal regions, together with the aforementioned migration from rural to urban areas have been the key factors in shaping the social structure of modern Turkey.

Also, and after decades of being known as a source country for substantial emigration, Turkey is now facing increasing challenges to its inward migration and asylum policies. Turkey has, in fact, long been a country of immigration and asylum. From 1923 to 1997, more than 1.6 million people migrated into Turkey, mostly from the Balkan countries. During the Cold War, thousands of asylum seekers fled to Turkey from the Communist states of Eastern Europe and the Soviet Union. The overwhelming majority of these migrants were recognized as refugees, and were resettled to third countries such as Canada and the United States by the United Nations High Commissioner for Refugees (UNHCR). In the late 1980s however, this pattern began to change, as increasing numbers of asylum seekers began to arrive from Iran and Iraq, as well as from other developing nations. In the years between 1988 and 1991 Turkey also experienced a mass influx of almost half a million (mostly Kurdish) refugees from Iraq, as well as large numbers of Albanians, Bosnian Muslims, Pomaks (Bulgarian-speaking Muslims), and Bulgarian Turks in 1989, 1992-1995, and 1999.

Today, officially sanctioned immigration into Turkey has to all intent and purposes become nonexistent. However, since the early 1990s Turkey has seen the arrival of a new form of "irregular immigration" involving nationals of neighboring countries, EU nationals, and transit migrants. Turkey allows nationals of Armenia, Azerbaijan, Georgia, Iran, Moldova, Ukraine, Russia, and the Central Asian republics to enter the country more-or-less freely - either without visas or with visas that can easily be obtained at airports and other entry points. A large number of these migrants are involved in small-scale trade. However, many subsequently overstay their visas and enter the 'shadow economy' working as domestic workers, commercial sex workers, or general labourers, especially on construction sites and in the tourism sector (the parallel with what is happening in Spain is most striking).

It is very difficult to give a reasonable estimate of the numbers of such irregular immigrants in Turkey. However, figures ranging from 150,000 to one million are often cited. To these groups must be added the victims of trafficking, particularly women. These are migrants who have either been coerced or deceived into traveling to Turkey for commercial sex work, and remain in Turkey against their wishes. There is also an increasing number of EU member-state nationals engaged in professional activities who are settling in Turkey, particularly in Istanbul, as well as European retirees in some of the Mediterranean resorts. They, too, constitute a relatively new phenomenon in terms of immigration into Turkey, and their numbers are estimated to be in the 100,000-120,000 region.

This situation marks a complete turnaround in Turkey's status as a migration country. As early as 1961 Turkey negotiated the first migration recruitment agreement with the then Bonn-based Federal Republic of Germany. Subsequent agreements with the Netherlands, Belgium, Austria, France, and Sweden were to follow during the years between 1964-67. At the present time about 3 million people of Turkish origin are estimated to live in Western Europe, making Turks the largest single migrant population in Europe. 4.8% of the Turkish population could be classified as European migrants. The largest Turkish population resides in Germany (2 million migrant Turks). In the 60s, 70s and 80s the Turkish government supported emigration due to the high level of unemployment and the need for the economic support which came with it in the form of remittances. During the late 1990s, migration to Europe has been mostly relegated to the spouse-selection of Turks by exising Turkish emigrants, political motives (Kurds), and undocumented labour migration.

For a variety of reasons a good deal of Turkish migration has not been in the form of permanent emmigration. In part this has been a result of policies implemented by the German and other EU governments, and in part it reflects the aspirations of the Turks themselves. New Turkish emigrants outnumbered return migrants during 1973-81 and 1986-94. Migration patterns showed more return migrants than emigrants during 1982-85. The balance between emigrants and return migrants was 17,347 emigrants (to Germany) in 1994. During the 1990s, the most popular destinations for Turkish migrants were the Commonwealth of Independent States, followed by North African and then Persian Gulf countries. Most recent migrations are project-related among semi-skilled and skilled male workers.

Turkey is currently the 17th most populous country in the world and the median age of the population is 27.7 years. The Turkish population is projected to increase, though the rate of population growth is now set to slow over time. The UN envisions an average population growth rate of 1.1% a year in its medium-variant scenario. More importantly, the population is young and around 70% (and rising) of the population is of working age (15-65 years old). A more rapid increase in the population of working age compared with the dependent population should help boost private savings from the currently modest levels. Labour force growth is expected to average around 1-2% over the next decade. As was the case in Asia in 1970-904, this “demographic dividend” has been and will continue to be a considerable factor in boosting Turkish GDP growth.

Turkey is already a fairly open economy, both in terms of trade in goods & services and portfolio flows, especially when adjusted for the size of the economy. Imports and exports combined represent more than 60% of GDP (and rising). Export growth has been phenomenal over the past decade with exports of goods rising from around 20% of GDP in 1994 to 30% in 2003. With continued economic reform and closer economic integration with the EU, the Turkish economy will become even more open in trade terms, which should benefit efficiency and growth. We expect exports to reach 40% of GDP by the end of the decade.

On the other hand the quality and level of education in Turkey has been modest when compared with other rapidly developing economies. However, Turkey has made sizeable progress in this area in recent years, especially as regards basic and intermediary skills. Primary school enrolment is close to 100%, which puts Turkey at a level comparable to Central and Eastern European countries. But at a level of 55-60%, secondary school enrolment ratios are still comparatively low. Only a gradual increase in the secondary school enrolment ratio is the most likely scenario given the government’s limited fiscal resources, although priorities may well change here. So there still remains plenty to be done. Improvements are modest and slow: while overall literacy levels are low, youth literacy levels are still only 'fair'. However, in stock and flow terms, even this modest improvement is indicative of future growth, and therefore a better indicator than the overall literacy levels. After all, it is these teenagers that will soon enter the labour market, steadily replacing as they do their on-average less well-educated parents and grand-parents. This being said, overall Turkey’s human capital endowment still compares unfavourably to other emerging market economies (see World Bank Education Index table). Additionally, Turkey has a very well-educated, often foreign-trained elite and a number of first-class universities churning out highly-skilled graduates.

To be edited and continued

Thailand

Life Expectancy and Model Limitations

Despite having achieved a certain degree of recognition and empirical successes, the variable rate-of-growth effect model is unlikely to provide a complete theoretical framework for understanding the behavior of saving rates and capital flows over whole course of the demographic transition. In particular, the variable rate-of growth effect model describes only one possible relation - that of a hypothetical steady-state relationship between dependency and saving rates - a shortcoming which in part from its life-cycle theory ancestry.

Continuing demographic change over the past half century, however, suggests that we look more to transitional dynamics if we want to more fully understand the observed correlation between dependency and saving rates and capital flows.

In this vein Higgins and Williamson (1996) show that the variable rate-of-growth effect model can be subsumed under the standard textbook neoclassical growth model, suitably modified to incorporate an overlapping generations population (see e.g., Blanchard and Fischer, 1988). The standard growth model is, in essence, simply an open-economy, steady-state version of the latter. The textbook overlapping generations model need only be augmented by adding a third period of life - childhood alogside old age — in order to accommodate a more comprehensive study of dependency effects.

References:

Bloom, D. and D. Canning, 2003a. From demographic lift to economic lift-off: the case of Egypt, Applied Population and Policy 2003:1(1) 15–24.

Bloom D, Canning D (2003b). Contraception and the Celtic tiger. Economic and Social Review, 34(3):229-247.

Coleman, D.A. The Demographic Transition in Ireland in International Context. In J.H. Goldthorpe and C.T. Whelan (eds.), The Development of Industrial Society in Ireland. Oxford: Oxford Press, 1992.

Fortin, Pierre, 2002, The Irish Economic Boom: Facts, Causes and Lessons, Industry Canada Research Publications Program, Ottawa, Ontario, Discussion Paper Number 12 May 2002

Kinugasa, T. (2004). Life Expectancy, Labor Force, and Saving, Ph.D. Dissertation. University of Hawaii, Manoa.

Kinugasa, T. and A. Mason, 2005, The Effects of Adult Longevity on Saving, mimeo, University of Hawaii, Manoa

LAL, Deepak, 1991, "World Savings and Growth in Developing Countries", Discussion Papers in Economics No. 91-05, University College, London.

Mason, A. 2005, Demographic Transition and Demographic Dividends in Developed and Developing Countries, Paper presented at the United Nations Expert Group Meeting on Social and Economic Implications of Changing Population Age Structures, United Nations Department of Economic and Social affairs, Mexico City, Mexico, September 2005

J. Murphy-Lawless and J. McCarthy (1997), Social Change; Public Policy; and Irish Fertility in the Twentieth Century. Manuscript.

Schultz, P.T. 2004, Demographic Determinants of Savings: Estimating and Interpreting the Aggregate Association in Asia, Economic Growth Centre, Discussion Paper No. 901, Yale university, New Haven.

Yavuz, Sutay, 2005, Fertility Transition and the Progression to Third Birth in Turkey, Max Planck Institute for Demographic Research, MPIDR Working Paper WP 2005-028 September 2005

Demographic Dividend

2006-06-03T05:07:00.000-07:00

Recent use of the expression demographic dividend seems to originate in a studyof East Asian growth carried out by David Bloom and Jeffrey Williamson (Bloom and Williamson 1998). In the study they used quantitative results obtained from cross-country econometric regressions with the objective of calculating the contribution made by age structure changes to the 'spurt' in East Asian economic growth.

Simply stated what has become known as the demographic dividend occurs when a fall birth rates following an initial mortality decline leads to changes in the age distribution of a society, with these changes meaning less collective expenditure is required to meet the needs of the youngest age groups (a reduction in downward intergenerational transfers in Lee's terminology: Lee, 2003) . In this way resources are released which may be used for investment in economic development and for improved family welfare. Essentially a falling birth rate makes for a smaller share of the population in the younger, dependent, ages and for relatively larger share in the adult working age groups who comprise the productive labor force. Thus the ratio of productive workers to child dependents in the population is improved, and with this the potential per capita income. In principal this should make for faster economic growth and fewer burdens on individual families and collective welfare systems.

Now according to Bloom and Williamson the 'demographic dividend' leads to opportunities for growth in output per capita for two principal reasons.

Firstly, there is a simple age-structure impact on total GDP due to a simple 'factor availability' or 'growth accounting' effect in that a rising share of the total population in the working-age group increases the ratio of producers to consumers. Obviously this situation is extremely favourable to the growth of output per capita. This 'composition effect' may be also be added to via the fertility decline which releases more women from childrearing activities and enables them to enter the labour market.

A second changing age-distribution impact is identified by Bloom and Williamson which they call the 'behavioural effects'. This behavioural impact takes a variety of forms. On the one hand a growing proportion of prime-age workers in the work force enhances overall productivity due to the well-know prime age worker productivity effect. On the other hand there are changes in aggregate saving and consumption following the life cycle pattern. This increase in saving can, in principal, make capital more available, and hence relatively cheaper. There may also be impacts from gender related issues associated with the fertility decline as changing attitudes to female emancipation leads to an increase in female participation in the education system, and hence a more educated female labour force The sum total impact of all of these is a further increase in potential growth in output per capita.

Clearly this demographic dividend is far from permanent. There is a limited window of opportunity. In time, the age distribution changes again, as the large adult population moves into the older, less-productive age brackets and is followed in its turn by the smaller cohorts that were born during the fertility decline. When this occurs, the dependency ratio rises again, as does the level of intergenerational transfers, but this time it is a question of the care and support needs of the elderly, rather than the demands imposed by the need to support a large young population.

The course of this process can be followed through what Bo Malmberg calls the Four Phases of the Demographic Transition schema.

Additionally, the dividend is not automatic. While demographic pressures are eased when fertility initially falls, some countries will take better advantage of this easing than others. Institutional structures matter. Some countries will act to capitalize on resources released and will use them effectively, while others will not. From the list of median ages, it can be seen that countries like Russia, Cuba and South Korea have passed through the demographic dividend epoch without gaining anything like the growth boost they should have. Iran seems to be about to add its name to this list.

This is important since, in time, the window of opportunity closes, and those countries that who have not found the way to take ample advantage of the dividend will find it hard to make the necessary leap when the winds of demography are finally blowing against them.

The demographic dividend is conventionally thought to be delivered through the operation of a number of interconnected mechanisms.

Labor Supply

As the demographic transition follows its course the generations of children born during the high fertility years enter adult life and become workers. Women who are now having fewer children than before are released from childrearing responsibilities and are able to take jobs outside of the home; also, as the transition moves forward, and years of compulsory education increase, younger women tend to become better educated than those to be found in the older cohorts, and are thus more productive once inside the labor force.

Savings

Mature working-age adults tend to earn more and thus can potentially save more than new-entrants to the labour market, or those setting up an independent home for the first time. Thus the larger generations who work their way through the labour force as the age pyramid changes favour greater personal and national saving. This ability to save becomes even greater as the 'thick cohorts' move into their 40s, especially as in the first instance the generation-span is smaller, and their own children rapidly become wage-earners themselves and hence soon require less support. Thus personal savings continue to grow and are able to serve as a source of investment funds. Countries steadily move from being heavily dependent on external finance, to a position of relative financial self-sufficiency.

Human Capital

Having fewer children normally enhances the health of both mother and child. Female participation in the labor force, in turn, enhances the social status and personal and financial independence of women. Also fewer children normally means fewer and better educated ones. More investment is allocated to each individual child.

There may however be other mechanisms at work. The most important among these is undoubtedly health. It is long been known that the process of economic development is associated with an improvement in the general level of health of the population. Conventionally this has been associated with an improvement in the nutritional environment and with a reduction of the disease load on young children.

One clear insight into how this process comes from the field of Life History Theory. Ronald Lee puts his finger on what is an extraordinarily important idea as follows:

"Turning to juvenile mortality when there is parental investment, contrast the death of a baby bird just after hatching with a death just before a baby fledges. The classic theory predicts equal selection against mortality at the two ages. However, the later death would be a total loss, whereas the early death would free up parental resources for greater investment in the surviving chicks, boosting their survival, size, and reproductive fitness, thus offsetting the direct effect and perhaps even increasing the survivors to maturity."

References

Bloom, D. E. and J. G. Williamson, 1998. Demographic Transitions and Economic Miracles in Emerging Asia. World Bank Economic Review vol.12, No.3, pp. 419-56.

Lee, Ronald, D. 2005. Rethinking the evolutionary theory of aging: Transfers, not births, shape senescence in social species, PNAS, vol. 100, no. 16, 9637-9642

Nasty Brutish and Short? III

2005-10-15T06:08:00.000-07:00

As I say above, following the initial mortality decline all societies are effectively ageing, the ageing is continuous, and at the present time it is hard to identify a natural barrier to this process. In this sense the transition doesn't really seem to have an 'end state', and thus can hardly be called a transition, since the word transition seems to imply something. If there is in fact a transition it is one from a society homeostatically balanced around high mortality to one which is pivoted around low and steadily declining mortality.

Having said this, and in fairness to Lee, what may be meant by ageing is a society with a comparatively high proportion of dependent elderly. On this view the initial mortality decline creates a dependency ratio which is considerably higher than that in the earlier agricultural society. This 'imbalance' takes many years to correct as fertility rates remain high and societies slowly recover the earlier ratios. But equilibrium is not recovered, and dependency ratios once more start to rise, this time amongst the elderly population. So this is what many may mean by ageing societies: societies where elderly dependency ratios rise (and continue to increase) above a certain notional level.

This way of looking at things has a certain validity, but it does beg one very important - indeed possibly critical from a policy perspective - question: just what do we mean by 'old'. The expression, like the terms modern and post-modern is a deceptive one, since it gives the impression of veing carved eternally in time, when in fact it is, of course, an extraordinarily relative one. To give one illustrative example, one populist Turkish politician got himself elected on a promise to introduce male pensions from the age of 43 and female ones from the age of 39 (something which, of course, resulted in the worst pension's crisis in history). He presumeably thought that 43 was 'old' and those who voted him into power evidently agreed. What we consider to be old is a socially defined (and hence relative) concept. It will hold different values at different times, and as life expectancy reaches ever higher limits we can expect our definition to adjust accordingly. This topic however, will have to await a later stage in the argument to receive the elaboration which it deserves. Simply consider this a foretaste.

Whatever the ultimate verdict on the validity of the phases schema, it should be noted that societies which enter the transition later tend to pass through it at an ever increasing rate. This if we take the mortality decline component we can see that gains in life expectancy have occured in the twentieth century in developing countries at rates which are rapidly by historical standards. In India, life expectancy rose from around 24 years in 1920 to 62 years today (a gain of 0.48 years per calendar year over 80 years), while in China, life expectancy rose from 41 in 1950–1955 to 70 in 1995–1999, (a gain of 0.65 years per year over 45 years.(Lee 2003) Fertility transitions since World War II have typically been more rapid than those for the developed countries, with fertility reaching replacement in 20 to 30 years after onset for those countries that have now completed the transition. Fertility transitions in east Asia have been particularly early and rapid, while those in south Asia and Latin America have been slower in starting but now seem to be accelerating (Casterline, 2001, United Nations Population Division, 2003).