This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: This chapter describes a formal demographic perspective on human
population dynamics. It ﬁrst attempts to summarise the way in which
human population dynamics are treated in the more technical
and theoretical demographic literature. The next section considers
some demographic fundamentals, including population structure
(especially the age and sex composition) and the three components of
population change: fertility mortality and migration. The third part
looks at some of the formal models which demographers have de
veloped to help understand population change. These models make
E several assumptions in order to simplify a complex reality. One ofthese
is that migration is zero: populations with zero migration are said to
be closed. An attraction of this is that, if migration can be ignored,
simple relationships exist between fertility, rationality, the population
: growth rate and the age structure. t In the {Ourth section population dynamics in the short to medium
term are considered. The age and sex structure of a population is
itself a dynamic feature, containing a record of the population’s past
fertility, mortality and migration. Moreover, the future age—sex struo
ture is determined by past and current events. Discussion of these
E aspects of population dynamics leads naturally in the ﬁfth section
to a consideration of population momentum, or what is sometimes,
inaccurately, called the ‘demographic time bomb’. The origins of
population momentum are explained. 2' Finally, in the sixth and seventh sections long—run population dy
namics are explored in the context of the demographic transitions in lo HELEN nacssrn AND PAUL COLLINSON E334" Hum” PJiUI “AM ‘0 ‘1 ‘U Mi g ‘ C P055 997361} 1’5",”
i pensﬁacaEw: . ll Q2 mioé’eé 5...! Val (M45
References E 2 car«(0P3 4&ﬁ 06L): iv) ﬂW , 100.2.
Clarke, ji. (1972). Geographical inﬂuences upon the size, distribution and E ' '
growth of human populations. In The Structure nguman Populations, ed. GHA E" DEWOgqu [126 p 67516. 6612.063 07?, [1315an
Harrison and AJ. Boyce, pp. 17—31. Oxford: Oxford University Press. E; POPZJZCZZZOW, dynamics
Harrison, GA. and Boyce, AJ. (ed) (1972). The Structure quumon Populations. ET
Oxford: Oxford University Press. E
Malthus, TR (1803). An Essay on the Princzjole of Population. {Volume 2 of The Works E"
of Thomas Robert Malthas, ed. EA. Wrigley and D. Souden (1986). London: E. ANDREW HINDE
William Pickering] E
E5 .
E: Introduction
E. .=. 17 1C5 ANDREW HINDE Europe and contemporary Africa. The relationships between fer—
tiiity, mortality and population growth enable closed populations to
be placed in fertility—mortality space’. Placing historical and con~
temporary populations in this space reinforces the idea that long—run population growth is naturally very slow. It also allows comparisons
to be made of the demographic transition in different cultures. Some demographic fundamentals Demography is about analysing the growth (or decline} of human
populations and changes in their structure. We cannot analyse pop
ulation growth without some straightforward and unambiguous way
of enumerating the members of a population, and this can only ex:
ist if we can work out who is, and who is not, to be included in the
enumeration. To demographers, a population consists of any group of persons who
can be delimited on the basis of some observable characteristic. The
most common way is to deﬁne the population on the basis of resi—
dence within a given geographical area. Thus we speak, for example,
of the “population of England’ as being all those persons normally
resident in England. However, residence is not the only criterion. We
could delimit populations on the basis of tribal afﬁliation: the Luo
population of Kenya, for example, using both geographical residence
and membership of a particular tribe as criteria. Necessary features
of characteristics used to delimit a population are that they be observ~
able and well deﬁned, so that we can use them to say whether any
individual person is or is not a member of the population at a given
time. Notice that although these characteristics are used to distinguish
separate populations, they themselves are inclusive: they are shared
features which bind the members of a distinct population together. In order to be able to analyse population growth and change, it is
essentiai to be able to identify the processes by which persons enter
or leave populations. It is fortunate for demographers that popula“
tions only change in size because of a limited, countable, range of
events. Consider, for example, the population of a particular country
at some time, say ijanuary 2000, which we might call P2000. Then the Demographic perspectives i9 population of that country on ljanuary 2091, P2001, is equal to P2999
plus the number of births during the year 2000, minus the number
of deaths, pins the number of people who migrate into the country
during the year, minus the number of persons who migrate out. The
difference between the number of births and the number of deaths
is known as natural increase (or decrease if deaths exceed births) and the
difference between the number of immigrants and the number of
emigrants is known as net migration. Analysing changes in the size of human populations, therefore, in~
volves the analysis of the processes by which births, deaths and migra~
tion events come about. The process which produces births is known
as fertility; the corresponding process which results in deaths is called
mandolin. The three processes of fertility, mortality and migration are
known as the components of population change. Demographers are interested particularly in the intensity with
which these events occur in a particular population. Since a large
population will tend to generate more events than a small popula
tion, the absolute numbers of, say, births and deaths in a particular
time period are of limited use as measures of this intensity. Therefore,
demographers use what are called rates. A demographic rate is a ratio
of events of a particular type, for example deaths, to the number of
persons exposed to the risk of experiencing that type of event. Thus
the crude death rate is equal to the number of deaths in a given period
divided by the ‘average’ population during that period. it is normally
important that the events and the population ‘exposed to risk’ cor—
respond. That is, we need to make sure that the persons exposed to
risk really are at risk of experiencing the event in question, and that
we do not include events in the numerator which do not occur to
persons in the denominator. There are some exceptions to this, one
being the crude birth rate (number of births in a given period divided
by the ‘average’ population during that period) in which men, who
do not give birth, are included in the denominator. Because peopie are not all identical, all human popuiations have a
structure, By population structure demographers mean the distribution of
various characteristics across the members of a population. The char—
acteristics most commonly considered by demographers are sex and
age. Certainly, these will be the most important variables so far as this 20 ANDREW HINDE chapter is concerned. Other variables by which a population struc—
ture may be deﬁned include genetic make—up, socioeconomic class
iﬁcation, occupation, educational attainment, place of birth, ethnic
aﬁﬁiiation, etc. A very important feature of populations is that the intensity of the
components of population change varies according to people’s char—
acteristics. Thus, different subgroups within a population will have
different risks of experiencing births, deaths and migration. People
aged over 60 years have a higher risk of dying than do teenagers,
for example. For this reason, demographers tend to work with rates
speciﬁc to particular subgroups. The most commonly used of these
are agespeciﬁc rates, but in principle, rates can be calculated speci«
ﬁc to any subgroup of interest, for example, as deﬁned by variables
discussed above in regard to structure. Formal demographic models of population change If migration is ignored (that is, if we assume a closed popuiation),
simple relationships exist between a population’s fertility, its mortal—
ity and its rate of growth, and between these three variables and its
age structure. These relationships can be described mathematically,
and this permits the construction of elegant demographic models
of human population dynamics. Two sets of relationships are of
great interest in a closed population (Figure 2.2). The ﬁrst is that
linking fertility and mortality to the rate of growth. The second is
that linking the rate of growth and mortality to the population’s age
structure. Let us consider the ﬁrst ofthese. A convenient way to understand
how the relationship works is to imagine the population of a remote
island. Suppose that on this island there are 100 men and ISO women,
and suppose that each man is ‘married’ to one woman so that we have
100 couples. By ‘married’ we simply mean that the couple is in a more
or less stable sexual relationship, whether this is iegally formalised or
not. One plausibie measure of the rate of population growth is the
total number of children that these 100 couples produce. Clearly,
if this is greater than 200, then the next generation will be larger Demographic perspectives 21 5ertility Mortality Age
structure Figure 2.1. Relationships in a closed population. than the present generation, and the population will grow. If it is less
than 200, the population will decline. The total number of children
produced by the iOO couples can be divided by 100 to give the average
completed family size. It" the average completed family size exceeds
2, then the population will grow. Demographers often refer to the
average completed family size as the totalferiiligz rate (TFR). If, then,
on our remote island the TFR is 4:, each woman will produce four
chiidren. Suppose that half of these are boys and half girls. The result
will be that our 300 couples will have 400 children, 200 sons and 200
daughters. If the same TFR is maintained for subsequent generations,
then 800 grandchildren will be born, i,600 great—grandchildren, and
so on. The population will therefore double in size every generation.
This, of course, is what Thomas Malthus meant when, in his Essay) on
the Principle if Population (1 798), he described popuiations as increasing
in a :geometrical ratio’. The TFR is, in fact, the sum of a set of agemspeczﬁc fertility rates
(ASFRs). Recall the deﬁnition of a demographic rate in the previ
ous section. The ASFR is deﬁned as the number of births to women
of a given age in a given year divided by the number of women of
that age, i.e. events divided by those exposed to risk. It measures the
number of births that the ‘average’ woman can expect to have in a
year while she is at that age. So, for example, the ASFR for a woman
aged 20 years last birthday is a measure of the number of children the 4!. ANDREW HINDE average woman has in the year between her 20th and 21st birthdays.
In practice, this will be less than i in almost all human populations, as
most women in most populations will have no children at all between
their 20th and Qist birthdays. If we sum the ASFRs at all ages from
the youngest age at which women hear children to the oldest age, the
result will be an estimate of the number of children which the average
woman will produce during the whole of her childbearing period, or,
in other words, the TFR. Population growth, therefore, depends on fertility. However, it mat~
ters whether the children are boys or girls. To see this, imagine the
populations on three remote islands, A, B and C. Each of these
islands contains a population of 100 couples. Suppose that on all these
islands each couple has exactly four children, i.e. the "FF R is equal
to 4. On island A equal numbers of boys and girls are born. However,
on island B three boys are born for every girl, and on island C thred
girls are born for every boy. Island A exhibits the doubling every
generation that we have already described. On island B the original
100 couples produce 300 sons and 100 daughters, whereas the 100
couples on island C produce 100 sons and 300 daughters. Thus in
the first generation, the population ofboth islands doubles. However,
in the next generation things become more interesting. Assuming
that the women continue to have four children each, then on island
B only 400 children will be born in the third generation, whereas
on island C, each of the 300 daughters of the ﬁrst generation will
produce four children, making a total of £200 grandchildren (300
grandsons and 900 granddaughters). Oi” course, achieving this would
involve polygyny or monogamy with extra—marital childbearing. For
the population of island B to match the growth of island C, each of
the daughters of the first generation would have to produce i2 child
ren. Even supposing that polyandry were widespread, as there are,
after all, three males for every female in the second generation, it is
most unlikely that this will he achieved. Indeed, the highest reliably
recorded fertility in a human population is around ten children per
woman among the Canadian Hutterite population during the 1920s
and 1930s. The key point to take away from this stylised example is that the rate of population growth depends on the number of girls
born rather than on the number of boys. Demographic perspectives 2 3 In practice the sex ratio of births among human populations varies
rather little; it averages 105 or 106 boys per 100 girls. This implies that
a TFR ofi will leadto a population not quite doubling itself every gen
eration. It is likely that an awareness that, so far as population growth is
concerned, it is daughters that matter, almost certainly iay behind the
female infanticide practised by certain tribal populations in the past. This mention of sex—selective infanticide (see Rousham and
Humphrey, Chapter 7) leads neatly into the ﬁnal factor determining
the rate of population growth in a closed population. This, of course,
is mortality. And what matters here is the chance that a daughter will
survive long enough to have children herself. Returning to the remote
islands, let us suppose that on islands D and E the sex ratio of births
is 105 boys per 100 girls. Suppose, however, that on island l) four
in every live girls horn survive to reproductive age, but that on island
E only one in every two girls does this. On island D, the l00 women
in the original generation will bear ti00 children, of which 395 will be
daughters (400 X 100/205). Only four—ﬁfths of these (i 95 x 0.8 2 156)
survive to reproductive age. These 156 daughters will hear 624:
(156X4) children in total, of which 304 will be girls {524x 100/ 205).
Of these, 80% will survive to reproductive age, resulting in a popu
lation of 243 granddaughters. The female infant and child mortality
has reduced the rate of population growth substantially This is even
more strikingly illustrated by island E, in which only half of those
born survive to reproductive age. It can be shown that the population
of island E will decline over the generations and eventually die out. For population growth, the mortality of females matters more than
that of males, though male mortality is not completely irrelevant.
Moreover, even for females, it is only mortality up to and during
the childbearing years that is of interest. Thus, to analyse population
growth, we need to measure agespeciﬁc female mortality up to about
age 50 years. Age5,0953% death rates (ASDRS) can be caiculated in a way
similar to that of ASFRS, by dividing the deaths in a given year to
persons of a speciﬁc age by the population of that age. A set ofASDRs
provides a complete description of the mortality experience of a pop»
ulation. In particular, ASDRs can be used to draw up a table of the
probability that a person will survive to at least a given age. This table
is known as a life table, and an example is shown in Table 2.1, using 24 ANDREW HINDE Table 2.1. Conditional prbbabz’lia'es qfdeatfz and
prababiiz‘gr @‘suraiaal to dg'ﬁrerent ages,females in
England and Miles, 1901—] 0 Conditional probability Probability of of death before next survival to at Age 9: speciﬁed age least age 2: 0 0.1 174 1.0000 1 0.0689 0.8826 5 0.0173 0.8218
10 0.0206 0.8076
15 0.0143 0.7990
20 0.0173 0.7876
25 0.0208 0.7740
30 0.0265 0.7579
35 0.0334 0.7378
40 0.0414 0.7131
45 0.0529 0.6836
50 0.0705 0.6475
55 0.1001 0.6018
60 0.1374 0.5416
65 0.1942 0.4672
70 0.2983 0.3764
75 0.4116 0.2641
80 0.5436 0.1554
85 1.0000 0.0709 M
The conditional probabiiity of dying before the next
speciﬁed age is the chance that a woman will die prior to
the next speciﬁed age given that she is still alive at age x.
80, for example, the chance that a woman aged exactly
40 years will die before her 45th birthday is 0.0414
(just over 4%). Clearly, this conditional probability for
the oldest speciﬁed age x must be 1.0000, as everyone eventually dies.
Source: Woods and Hinde (1987: 33). data which relate to the female population of England and Wales
in the decade 1901—40. Under this mortality regime, for example, a
woman had about an 80% chance (probability 0.7990) of surviving
to her 15th birthday, and about a 65 % chance (probability 0.6475) Demographic perspectives 25 of surviving to her 50th birthday. Associated with the probabilities of
survival to given ages are conditional probabilities of death within a
particular age range. These conditional probabilities are expressions
of the chance that someone Wili die within an age range given that
they are alive at the start of that age range. An important special case
is the infant mortality rate, which is the probability that a baby will die
before his or her ﬁrst birthday From Table 2.1, it can be seen that the
infant mortality rate for girls in England and Wales in i901—10 was
0.1l74. in other words, almost 12% of girls born during this period
did not survive until their ﬁrst birthday. Although the mortality of
males is not so important for understanding population growth, it
is worth pointing out that in most human populations, the mortality
of males exceeds that of females at all ages. Infant and child mortality
is discussed further by Rousham and Humphrey (Chapter 7). The stylised example described in this section shows that popula—
tion growth depends on three things: the average number of children
born to women, the sex ratio of births and the chance of a female child
surviving to reproductive age. One measure of population growth
which incorporates all of these is the net reproduction rate (NRR). The
NRR is approximately equal to the TFR muitiplied by the propor—
tion of births that are girls multiplied by the probability of a woman
surviving to the mean age at childbearing. The mean age at child«
bearing varies a little between populations, but is between 27.5 and
30 years in most cases. Therefore, suppose that a population had a
TFR of 4.0, that i05 boys were born for every 100 girls, and that
the mortality of females was described by Table 2.1. The chance that
a woman will still be alive at exact age 25 years is 0.7740, and at
age 30 years it is 0.7579. So the probability of survival to the mean
age at childbearing (27.5430 years) may be estimated at, say, 0.76.
The NRR is, then, calculated as 4X(100/205)><0.76 = 1.48. The
NRR measures the size of the next generation relative to the size of the present generation. 80 an NRR of 1.48 means that the next
generation will be 48% larger than the current one. The second important relationship is that between the rate of
growth, the mortality and the age structure of a population. The
number of persons alive at any age x, at a particular time, in a closed
population, is the product of two factors: the number born at years ago ZO ANDREW HINDE and the probability o...
View
Full Document
 Spring '11
 Marsh

Click to edit the document details