As I’ve mentioned in previous posts, in many developed countries the level of reproduction, as measured by the Total Fertility Rate, is currently below the replacement rate (the level necessary to sustain the population). With current levels of infant mortality, the replacement rate is only about 2.1, but the TFR in many countries is below this. For example:
Australia…………………1.8
Canada………………….1.6
France……………………1.7
Germany………………..1.4
Ireland……………………1.9
Italy………………………..1.2
Japan…………………….1.4
UK…………………………1.7
USA……………………….2.1
(Estimates for 2002, source ‘The World Fact Book’)
With the exception of the USA the TFRs are all below the replacement rate. American readers may take national pride in their greater fertility, but they should note that the TFR for non-Hispanic whites is below the replacement rate, and it is only the high rate among Hispanics that pulls up the US average.
If these TFRs continue indefinitely, then in the absence of net immigration total population would eventually fall. This point has not yet been reached (except perhaps in Italy), because in most developed countries there is significant net immigration, and life expectancy is also still rising. In the UK, for example, population is expected to grow from its present 60 million to 65 million by 2050.
A declining population is not necessarily a bad thing. A rising population may be good for economic growth, but not for the quality of life. A lot depends on the density of population. The prospect of a declining population will be more attractive in Japan, with 336 people per square km., or the UK, with 244, than in the USA, with only 29.
However, my main concern in this post is whether we can expect the current low fertility rates to continue.
First, as I’ve mentioned previously, the TFR is an artificial construct which mixes up the current fertility of different cohorts of women. If there is a trend for women to postpone childbearing until later in their reproductive lifespan, then there will be a time-lag during which the TFR will systematically understate eventual cohort fertility. In the UK, surveys of women show that they intend on average to have around 2 children, though the current TFR is only 1.7. Longitudinal studies show that women on average get close to their intended family size, though many women who postpone childbearing find it more difficult to conceive than they had expected, as fertility declines from age 30 onwards. I would guess that over the next decade the TFR will increase somewhat from the current low levels, but not quite to the replacement rate.
Over the longer term, can we expect the TFR to return to the replacement rate or even higher? Now that reliable contraception is available, the fertility of women depends largely on how many children they and their partners want. Assuming that the biological impulse to reproduce is constant, the demand for children will depend on the other costs and benefits of having them. Demographers and economists have analysed these non-biological factors. In developed countries the economic motive of having children to put them to work, or to support their parents in old age, is no longer relevant. The economic factors are therefore largely costs rather than benefits, such as the costs of food, housing, education, health, and child-care, and the opportunity cost for women of not working during pregnancy and after childbirth. In most countries the costs are to some extent spread over the whole community through taxes (e.g. to pay for education), but there is probably no developed country where it is economically advantageous to have children.
It is unlikely that the costs of children will fall greatly. A reduction in population might reduce the cost of housing, which is a major burden for young families in some countries, but this is a very long term prospect. It is also possible that governments will provide financial incentives for having children, but such policies have not proved effective in the past (notably in France), and in any event resources will be stretched to provide for the increasing number of the elderly.
Purely on economic grounds it is therefore difficult to foresee any major increase in fertility. But all of this assumes that the underlying impulse to have children is constant. This is doubtful. From a biological point of view, fertility is subject to natural selection. We expect average fertility to rise until it reaches an optimum, where the number of offspring in a family who themselve survive and reproduce is at its maximum. If a family has more offspring than the optimum, then the higher number of children will be more than offset by their lower rates of survival and/or reproduction – the parents will have more children but fewer grandchildren than the average.
The optimum fertility for a species depends on its ecological conditions and way of life. There is a familiar distinction between r-selected species, where a large number of offspring are produced, but most of them die, and k-selected species, where a small number of offspring are produced, but a high parental investment in each ensures that a high proportion survive. However, even in a k-selected species the optimum level of fertility will not in general be the same as the level that minimises the mortality of offspring.
Man is a highly k-selected species. Optimum fertility is unlikely to be vastly higher than the minimum number necessary to replace the parents. However, it seems clear that the selective optimum must be higher than present TFRs, and somewhat higher than the present replacement rate of 2.1. If average fertility were at the optimum, then parents with more children than average would have relatively few grandchildren. I don’t know of any recent figures on this point, but it seems highly likely that parents of 3 or 4 children have more grandchildren than parents of 2 children, contrary to the hypothesis that the optimum is around 2.
There is also a theoretical reason for expecting current fertility to be below the selective optimum. The reason is that the pattern of fertility has not yet recovered from a recent downward shock. Effective oral contraceptives only became widely available in the 1960s (and more recently in some Catholic countries), which is scarcely a generation ago. For the first time in history people who want to have sex but not children can reliably do so. The result has been a large increase in the proportion of couples having no children, or only one child.
But we would expect the pattern of fertility eventually to adjust from this shock. The mechanism for adjustment is that when average fertility is below the selective optimum, then those individuals who have more than the average number of children will contribute more to subsequent generations – they will have not only more children but more grandchildren. And if the propensity to have more children has positive heritability, then this propensity will be passed on, and the average fertility of the population will rise.
To illustrate this with a crude example, suppose that out of every 10 women in the present generation, 3 have no children, 3 have one child, 2 have 2 children, 1 has 3 children, and 1 has 4 children. The total number of children is 14, and the average per woman is 1.4 (which is within the current range of TFRs in developed countries). Now suppose that heritability is 100%, so that every female has as many children as her own mother. On these assumptions the average number of children born to each female of the second generation will be 2.57. Average fertility has almost doubled in a generation. This largely reflects the fact that in this model the most fertile 20% of women contribute half of all the children in the second generation (7 out of every 14), and we
assume that the daughters from these families inherit their mothers’ high fertility.
Heritability of fertility is very unlikely to be 100%. However, it is also unlikely to be zero. With reliable contraception, fertility becomes largely a matter of psychology. Since factors of personality, however defined, are usually about 50% heritable, it is reasonable to suppose that the psychological factors influencing fertility are also heritable. (Even if the basis of heritability were environmental rather than genetic, fertility would still be transmitted for several generations, until the environment of the descendants had regressed entirely to the mean.) One thing we can say for certain is that women who have no children will have no grandchildren. So in an age of contraception any heritable factors leading women to be averse or indifferent to children will rapidly be eliminated from the population.
I conclude that over the course of another 2 or 3 generations, fertility is likely to rise back above the replacement rate, and populations in developed countries will grow again. If infant mortality does not also rise, at some point pressure on resources will lead governments to introduce measures to discourage fertility. Of course, over this time scale a lot of other things will have happened, and rising fertility may be the least of our worries.
I should mention that the main argument here is not new. It was stated many years ago by the physicist Sir Charles Galton Darwin (a grandson of you-know-who), but it has been generally neglected.
DAVID BURBRIDGE
Comments are closed.