STA 2110 FUNDAMENTALS OF VITAL STATISTICS (45 Contact

Hours)
Pre-Requisites STA 2100 Probability & Statistics I

(a) Course Purpose

This course introduces the basic techniques of demographic analysis and
interpretation of vital statistics.

(b) Learning outcomes

By the end of this course the student should be able to;

(1) Identify and compare the advantages and disadvantages of the dif-
ferent sources of vital statistics data.

(2) Describe basic demographic indicators and elaborate on their com-

putation and interpretation.

(3) Introduce population projection calculations and analysis

(4) Define and differentiate the demographic concepts, terminology and

formulas.

(5) Construct a Lexis diagram.

(6) Construct and interpret simple life tables;

(7) Perform direct and indirect methods of standardization.

(8) Describe the relations and calculate indicators in a stationary pop-

ulation

(9) Derive the mathematical relationships in a cohort life table.

(10) Estimate the rate of change in a population

1
 (11) Project a population using appropriate equations and assumptions.

(c) Course Description

Sources of data available for vital statistics. Population composition
and change measures. Measures of mortality, fertility, marriage and mi-
gration levels and patterns ; period and cohort methods of description
and analysis Life table, population standardization ;direct and indirect
methods; Constructing a lexis diagram, the determinants of age structure
and the intrinsic growth rate; use of survey data; the interpretation of
demographic statistics; population dynamics and population projection
techniques.

(d) Teaching Methodology

Lectures, Tutorials, Self Reading, Discussions and Student Presentations

(e) Instructional Material and Equipment

Black or White Boards, Chalk or White Board Markers, Dusters, Com-
puter and Projector

(f) Course Assessment

End of Semester Examination (70%); Continuous Assessment Tests (20%);
Assignments (10%)

Course Text Books

[1] Mckenzie S. Vital statistics, Churchill Livingstone ; ISBN 13:978 -

0729541497.2013.

2
[2] Poston, D and L Bouvier Population and society: An introduction
to demography. Oxford. HB 849.4 POS. 2010.

[3] Brown R. L.,Introduction to the Mathematics of Demography,2nd

edition, ACTEX Publications, ISBN: 9780936031149, 1993.

[4] Kpedekpo, G. M. K. Essentials of Demographic Analysis for Africa.

Heinemann. ISBN 0 435 97390 8, 1982.

Course Journals

[1] Lifetime Data Analysis (Lifetime Data Anal.) [1380-7870; 1572-9249]

[2] The Annals of Statistics (Ann. Stat.) [0090-5364]

[3] Statistics (Statistics) ISSN[0233-1888]

Reference Text Books

[1] Hinde A,Demographic Methods, Arnold publishers, ISBN:

9780340718926, 1998.

[2] Lee, E.T., Wang J., Statistical Methods for Survival Data Analy-
sis,3rd edition, John Wiley & Sons, ISBN: 9780471458555, 2003

[3] Rowland T. R., Demographic Methods and Concepts, OUP Oxford

publishers, ISBN: 9780198752639, 2003.

3
Reference Journals

[1] Annals of Statistics ISSN • 0090-5364

[2] Journal of the American Statistical Association ISSN • 0162-1459

(print) 1537-274X (web)

[3] Journal of the Royal Statistical Society (Series B) Online ISSN: 1467-
9868.

4
1 INTRODUCTION
Vital statistics provides an introduction to demography. The word ” demog-
raphy” is derived from the Greek words demos, people, and grapho, to write.
Demography can be defined as the study of human populations including
their composition, distributions, densities, growth and other characteristics
as well as the causes and consequences of changes in these factors. Demog-
raphy focuses its attention on three readily available human phenomena:

• Changes in population size (growth or decline)

• The composition of the population and

• The distribution of population in space.

Demography deals with major “demographic processes” namely fertility,

mortality and migration. These processes are continually at work within a
population determining its size, composition and distribution.

1.1 Uses of demography

Much sociology appears ethereal and difficult to grasp. Demography is one

of the more pragmatic areas of Sociology.

1. Demography and Business Business can put the study of population

to specific uses. If, for example, a business sells a product that is
desired by age-specific groups, that business can use the census to dis-
cover communities where members of that age-specific group live. For
example, a business may want to open an automobile dealership that

5
 specializes in expensive sports cars. Demographic awareness could help
the potential dealer in finding neighborhoods where young and middle-
aged affluent people live. Placing a Corvette shop in a working-class
community populated mainly by low-income middle-aged and elderly
people could be disastrous.

2. Demography and Government Demography is also an essential tool

used in ensuring equal representation in Parliament.

3. Allocation of Money for Social Services: This point also pertains to the
relationship between Demography and the Government (part B). Given
that it is important for the allocation of money to poor people, one
can understand why disadvantaged groups in society are at a farther
disadvantage when the Census Bureau under counts those populations.

1.2 Sources of demographic data

Demographic data are important in providing factual basis for decisions on

matters of public policy and action concerning social and economic affairs.
These data can be processed to indicate present and future requirements
of the population in terms of the types and extent of social needs of the
society such as health, education and employment. The major sources of
demographic data include:

• Census

• Registration of vital events (Records)

• Sample surveys

6
 • Ad-hoc Demographic studies

1.2.1 Census

Census is defined as an enumeration or complete population count at a point

in time within a specified geographical area. A census provides more reliable
and accurate data if properly enumerated. The conducting of censuses has a
long history, but modern censuses were first held in the 17th century. By the
end of eighteenth and beginning of the nineteenth century many countries
had held their first censuses. There are two techniques of conducting census,
dejure and defacto.

DE JURE

This technique is the counting of people according to the permanent place of

location or residence.

Advantages

• It gives permanent picture of a community.

• It provides more realistic and useful statistics

Disadvantages

• Some persons may be omitted from the count. A household member

who is temporarily away from home may be missed from being counted
unless the enumerator makes sure that nobody is missing.

• Some may be counted twice.

7
 • Information collected regarding persons away from home is often in-
complete or incorrect.

DE FACTO

This technique of conducting census refers to counting persons where they

are present at the time of the census period.

Advantages

• There is less chance for the omission of persons from the count.

Disadvantages

• Difficult to obtain information regarding persons in transit. These

are persons who are, for example travelling and have left their area of
permanent residence but haven’t reached the area of destination during
the census day.

• It provides incorrect picture of the population in a community.

• Vital statistics are usually distorted (in areas with high migration)..

A census is useful for: • Planning • Calculating health indicators and vital

indices.

Steps of coducting census

The major steps to be followed in a census include.

• Planning and preparation

8
 • Collecting information

• Compilation and analysis

• Evaluation

Qualities of a Census

A census must have the following qualities

• It should include every individual in the area (no omission or duplica-

tion).

• Information should relate to a well – defined point in time.

• It should be taken at regular intervals (preferably every 10 years).

• It should refer to people inhabiting a well defined territory.

• Information be obtained from personal contact

1.2.2 Registration of Vital events

It is a regular and continuous registration of vital events. Civil registration

system, which records births, deaths, marriages etc. (vital statistics), enables
rates of population growth to be calculated; but are much less adequate
than national censuses. In developing countries where illiteracy rates are
high and communications are poor, the problems of recording births and
deaths are immense not only in rural populations but also in urban areas.
However, efforts are being made to improve the collection of these data in
many countries.

9
1.2.3 Sample Surveys

A sample survey is another source of demographic data carried out in a scien-

tifically selected area which covers only a section or portion (sample) of the
population under consideration. As countries have began increasingly to for-
mulate population policies as part of their development programs, especially
since 1970; so they have required more data about their populations. De-
mographic surveys have become increasingly important as a comparatively
rapid method of obtaining such information. It was recognized that the qual-
ity of demographic data available for the developing world was poor and the
need for better data both for scientific study and for policy application is
also recognized.

1.3 Study Questions

1. What is Demography?

2. Name and explain the different sources of demographic data?

3. State the steps that should be fallowed to undertake a population cen-

sus.

4. Explain the advantages and limitations of sample surveys.

10
2 DEMOGRAPHIC MEASURMENT TOOLS AND TECH-
NIQUES
After completion of this chapter the students should be able to:

• State the measurements used in demographic data

• Calculate ratios, proportions and rates and discuss their importance

• List and calculate the different measures of fertility and mortality

The major demographic processes of fertility, mortality and migration con-

stitute the basic components to determine the size, composition and distribu-
tion of a population which require basic tools and techniques of measurement.
The tools of measurement include:

• Ratios

• Proportions and

• Rates

Ratio

Ratio is a quotient of any two demographic quantities. It is the result of

dividing one quantity by another. Ratio quantifies the magnitude of one
occurrence or condition in relation to another. It is expressed in the form of:

a/b

For example the sex ratio , the dependency ratio. Note

11
 • The numerator and denominator (a and b) are defined for a specific
geographic area and period of time.

• The numerator may or may not be a sub – group of the denominator.

Proportion

A proportion is a ratio which indicates the relation in magnitude of a part

of the whole. The numerator is always included in the denominator. A
proportion is usually expressed as a percentage.

Rate

A rate measures the occurrence of some particular event (example death)

in a population during a given time period. It is a statement of the risk
of developing a condition. It indicates the change in some event that takes
place in a population over a period of time. It is defined per unit of time. A
rate comprises the following elements:

• Numerator

• Denominator

• Time specification and

• Multiplier or constant (100, 1000, 10000, 100,000 etc)

Note A Rate is a ratio or a proportion in which the numerator is part of the

denominator. All rates are ratios, but all ratios are not rates.

12
2.1 Techniques of Demographic Measurement

A change in the overall size of a population is the result of three cumulative

changes in the number of births, deaths and migrants. Since births and
deaths occur continuously, and since people frequently change their place
of residence, they are the bases of demographic analysis. Depending on
their direction and magnitude these processes are important for social and
economic planning, in assessing the present needs and the needs of the future
expansion of socio-economic infrastructure. The techniques for measuring
fertility, mortality and migration are described further.

2.1.1 Measures of Fertility

Fertility is the reproductive performance of an individual, a couple, a group

or a population. It is meant the actual bearing of children. Some demogra-
phers prefer to use natality in place of fertility. Fertility leads to increase of
population. It differs from fecundity – which refers to the physiological ca-
pability of a woman to reproduce. A woman’s reproductive period is roughly
15 – 49 years of age. The most important measures of fertility include:

Crude Birth rate(CBR)

The crude birth rate indicates the number of live births (children born alive)
per 1000 mid – year population in a given year.

(Births/P opulation) ∗ 1000

13
In the world CBR varies widely from population to population. It is high
for population of the developing countries and low for those of the developed
ones.

Limitations

1. Only a crude estimate of fertility, all the population included in the

denominator is not ‘exposed’ to the risk of pregnancy.

2. Not good for comparing fertility across populations, as variations in

age distribution of the populations being compared will affect the birth
rate.

General Fertility Rate (GFR)

The General Fertility Rate is the number of live births per 1000 females aged
15-49 years (fertile age group) in a given year. The GFR in more sensitive
measure of fertility than the CBR, since it refers to the age and sex group
capable of giving birth (females 15-49 years of age). It eliminates distortions
that might arise due to different age and sex distributions among the total
population. The major limitation of GFR is that not all women in the
denominator are exposed to the risk of child birth.

(N o.of Births/N o.of f emalesagedbetween14 − 49) ∗ 1000

The GFR is a more refined measure than CBR to compare fertility across
populations. The GFR is approximately four times the CBR.

14
Age specific fertility rate

The Age specific fertility rate is defined as the number of children born alive
to females in a specific age group per 1000 females in that specific age group,
example (15-19), (20-24),….. (45-49) years of age.

ASF R = (N o.of livebirthstof emalesinaspecif icagegroupinayear)/(M id−yearpopulationof f e

This measure is Useful for

1. Comparisons in fertility behaviour at different ages.

2. For comparison of fertility at different ages over time.

3. for comparison of fertility across countries/populations.

Example: Consider the following data:- Find the age specific fertility rate,

Age. Population of female Live births ASFR

15-19 1611090 463631 288
20-24 1558276 427298 274
25-29 1425242 412878 290
30-34 1381174 380778 276
35-39 1632695 308671 189
40-44 1581373 281161 178
45-49 140055 239701 171
Total 10590405 2514118 1666

the general fertility rate

GF R = (2514118/10590405) ∗ 1000

= 237 per 1000

15
Total Fertility Rate (TFR)

The total Fertility Rate is the average number of children that would be born
to a woman throughout her life time or her child bearing age (14-49 years),
if she were to pass through all her child bearing years at the same rates as
the women now in each age group. The TFR sums up in a single number the
Age Specific Fertility Rates of all women at a given point in time. If 5 – year
age groups are used, the sum of the rates is multiplied by 5. This measure
gives the approximate magnitude of “completed family size”. The TFR is
one of most useful indicators of fertility, because it gives the best picture of
how many children women are having currently. In the previous example the
total fertilty rate is year.

T F R = 5(sumASF Rperwoman)

= 5[1666/1000]
= 5 x 1.666 = 8 Per woman The TFR is the best single measure to compare
fertility across
Exercise Find the general fertility rate and the total fertility rate for the
following data TFR = 5 x 1.276 = 6.38
i.e. 6.38 children per woman in her reproductive life.
Exercise
Consider the following data on the births and the female population in
a state for the year 2000 .Find the age specific fertility rate and the total
fertility rate Age group 2000 births female population
Solution

16
Age. Population of female Live births ASFR
15-19 1,237,721 117,583 0.095
20-24 978,136 268,987 0.275
25-29 979,623 283,111 0.289
30-34 989,693 380778 0.257
35-39 814,243 162,034 0.199
40-44 548,882 57,633 0.105
45-49 406,540 22,766 0.056
Total 5,954,838 1,166,465 1.276

Age group births Female population

10-14 300 165000
15-19 11,000 179000
20-24 20000 192000
25-29 22,000 222000
30-34 20,000 213000
35-39 10,000 212000
40-44 2000 210000
45-49 500 200000

Age group births Female population

10-14 300 165000 1.8
15-19 11,000 179000 61.5
20-24 20000 192000 104.2
25-29 22,000 222000 99.1
30-34 20,000 213000 93.9
35-39 10,000 212000 47.2
40-44 2000 210000 9.5
45-49 500 200000 2.5

17
 sum= 419.7
TFR = 419.7 X 5 = 2,098.5 live births per 1,000 female state residents in
2000 who live through their reproductive years.

Gross Reproduction Rate (GRR)

The Gross Reproduction Rate is the average number of daughters that would
be born to a woman throughout her lifetime or child bearing age (15-49
years), if she were to pass through all her child bearing age. This rate is
like the TFR except that it counts only daughters and literally measures
“reproduction”; a woman reproducing herself by having a daughter. The
GRR is calculated by multiplying the TFR by the proportion of female births
(Sex Ratio at birth).

Child – Woman Ratio (CWR)

Child woman ratio is defined as the number of children 0 – 4 years of age per
1000 women of child bearing age, (15 -49 years). This ratio is used where
birth registration statistics do not exist or are inadequate. It is estimated
through data derived from censuses.

2.1.2 Measures of Mortality (Death)

Mortality refers to deaths that occur within a population (reduction of pop-

ulation). The incidence of death can reveal much about the living standard,
the health status of a population and the availability of health services. Mor-
tality (Death) rates have three essential elements:

• A population group exposed to the risk of death (denominator)

18
 • The number of deaths occurring in that population group (numerator)

• A time period.

Measures of mortality include:

Crude Death (Mortality) Rate (CDR)

The crude death rate is the number of deaths per 1000 population in a given
year. As its name implies the CDR is not a sensitive measure (indicator) of
health status of a population. It is affected by particularly the age structure
of the population. Crude Death Rate also varies between populations of the
world.

Age specific Death (Mortality) Rates

Death Rates can be calculated for specific age groups, in order to compare
mortality at different ages. E.g. for infants (< one year of age), children 1-4
yeas of age, children under five years, etc.

ASM R = N umberof deathsinaspecif icagegroupX1000M id−yearpopulationof thesameagegro

Infant Mortality Rate (IMR)

Infant Mortality Rate is the number of deaths of infants under one year of
age (0-11 months of age) per 1000 live births in a given year. Infant (children
under one year of age) are at highest risk of death than any other age group.
The infant mortality rate is considered to be a sensitive indicator of the health
status of a community, because it reflects the socio-economic condition of the

19
population; i.e. the level of education, environmental sanitation, adequate
and safe water supply, communicable diseases, provision of health services
etc. These factors mostly affect infants and children under five years of age.
Hence, IMR widely varies between countries in the world.

Neonatal Mortality Rate (NNMR)

Neonatal period is the first month of age of an infant. Neonatal mortality

(death) is the death of infants under one month (<4 weeks). Per 1000 live
births. Neonatal mortality rate reflects mortality due to: . Maternal factors
during pregnancy . Birth injuries . Neonatal infection, etc. It is an indicator
of the level of prenatal and obstetric components of maternal and child health
care (MCH).

Post-Neonatal Mortality Rate (PNNMR)

The past neonatal age is the period of time between one month up to one
year. Post – Neonatal mortality (death) is deaths of infants one month (four
weeks) of age up to one year (1 – 11 months age of) per 1000 live births The
post-neonatal mortality rate reflects deaths due to factors related to;

• Environmental sanitation

• Infections (communicable diseases)

• Nutritional problems

• Child care etc.

It can be used as an indicator to evaluate Maternal and Child Health Care

services and socio-economic development of a community or country.

20
Maternal Mortaly (Death) Rate (MMR)

Maternal mortality rate is the number of maternal deaths related to preg-

nancy, child birth and post natal (peurperium) complications per 1000 live
births (usually per 100,000 L.B). It is a sensitive indicator of health status
of a population. It reflects the socio- economic status of a community.

Sex Specific Death Rates (SSDR)

Sex Specific Death Rate is the number of deaths among a specific sex group
(males or females) per 1000 population of the same sex group. Sex specific
mortality rate is used to determine which sex group is at higher risk of death
than the other.

Standardized death(mortality ) rate

In order to compare the mortality conditions prevailing in two different re-

gions (say A and B) or of the same region over two different time periods,
one has to take into account a number of factors such as age, sex, occu-
pation, race, marital status etc. Generally, it is difficult to gather such a
large amount of information which may not be of much use otherwise. Thus
to reach at a single index of mortality based on some kind of average of the
death rates for various segments of the population, the simplest measure that
one may use is CDR. However, the value of CDR is greatly affected by the
age and sex distribution of the population under consideration. For compar-
ing the mortality situation of two regions with altogether different age and
sex distributions, the results obtained by CDR will not be valid. Consider

21
the CDR of two regions A and B in terms of the age specific death rates:-
∑ A A
DA M P
CDR = A ∗ 1000 =
A
∑ XA X
P PX
∑ B B
DB M P
CDR = B ∗ 1000 =
B
∑ XB X
P PX
This shows CDR as the weighted arithmetic mean of the age specific death
rates, weights being equal to the proportion of population with corresponding
ages i.e.
P
∑X
PX
This implies that even if the age specific death rates in the two popula-
tions are comparable, their CDRs will differ as a consequence of the difference
in the age distributions of the two populations
To remove this defect we have to use the same set of weights in taking
the weighted average of the SDRs of two regions. The index thus obtained is
called standardized death rate or adjusted death rate. When age distribution
is taken as the basis of finding the standard set of weights, the standardized
death rate will be called age standardized death rate or age adjusted death
rate. Similarly, it may be adjusted for other characteristics such as sex, oc-
cupation etc. and similarly interpreted. A death rate may also be adjusted
for more than one factor simultaneously. There are two methods of stan-
dardization:
1. Direct Method of Standardization: In this method we take some other
population called a standard population. For computing age adjusted death
rate we take the population for different age groups in the standard popula-

22
tion as the set of weights.
Let PXS be the number of persons at age x in the standard population,
then standardized death rate or adjusted death rate for the two regions A
and B are given as: ∑ A S
M P
ST DR = ∑ X S X
A
PX
∑ B S
M P
ST DRB = ∑ X S X
PX
Thus the age-adjusted death rate is simply CDR that would be observed
in the standard population if it were subject to the age SDR of the given re-
gion. The standard population is generally taken to be the actual population
of a bigger region of which both A and B are subsets.
Thus standardized death rate is a good index for comparing the mortality
conditions of two regions. Any differences in their age specific death rates
will be faithfully reflected.

2.Indirect Method of Standardization: In the previous method we require

the knowledge of the number of persons and the SDRs for various age seg-
ments in the given region. Quite often we have the classification of population
by age but the SDRs for various age groups may not be available. However,
the total number of deaths in the given region may be known to us. In such
a case we use indirect method of standardization. This formula requires the
knowledge of the number of persons and SDRs for various age groups in the
standard population along with the values of PXA for the given region. Thus

23
the formula for indirect age adjusted death rate is given by
∑ S S∑ S
T otal number deaths in region A M P P
ST DR = ∑ A × ∑ XA XS ∑ XA
PX M X P X PX

Standardized death rate suffers from the drawback that the value of STDR
so obtained depends on the age and sex composition of the standard pop-
ulation chosen which may be quite different from that of the region under
consideration. In case there is not much variation in the age distribution of
the standard population from that of the region under consideration, there
is not much problem. However, if the age distributions of the two regions
under comparison differ significantly, it is better to take the population of
that region as standard whose death rate is of more concern to us. Also
choice of the standard region is quite subjective and may introduce bias in
the results. Indirect method of standardization is used whenever the requi-
site data for computing STDR by Direct method is not available. However,
the two methods would give exactly equivalent results if the SDRs of the
given region happen to be proportional to the SDRs of the standard region.
Example Which of the two locations, for which mortality data are given
below, is in your opinion healthier?

Age group location A Location B

Standard population population deaths local population Population deaths
Under 5 4,500 135 4000 144
5-15 10000 40 10500 63
25-65 12500 75 13500 81
above 65 3000 140 2000 102

24
2.1.3 Measures of population growth

The growth of the population of a region or country can be measured by :-

• crude rate of natural increase

• Vital index

• Gross reproduction rate

• Net reproduction rate

• Replacement index

Crude rate of natural increase

this is the simplest measure of population growth. It is obtained by subtract-

ing crude death rate and the crude birth rate. Its also known as the crude
rate of natural survival.
= CBR − CDR

also Annual crude rate of natural increase is given by

T otal births − T otal deaths

( )
M id year population

Example The following are figures obtained from a sample registration system
in a partcular city in the yea 2000. Mid year population=35272000
Total births=187236
Total deaths=72299

25
Find the crude rate of natural increase.

187236 − 72299
( ∗ 1000)
35272000

=3.26 per thousand

Vital Index

An index that takes into account the two most vital events , namely births
and deaths. The index is given by

T otal births
( )
T otal deaths

This index may be equal to 1 or less than 1 or greater than 1. A value of 1

indicates stagnation in population growth A value greater than 1 indicates
expected increase in the population while a value less than 1 indicates a
decline in the population.

Gross Reproduction rate (GRR)

Gross reproduction rate is a measure of the population growth. Along with

considering the age and sex composition of the population it also considers
the sex of the newly born baby as it is a female birth, which is ultimately
going to be responsible for growth in population. Thus Gross Reproduction
Rate may be defined as the rate of replenishment of the female population.
It is a hypothetical figure. It gives the number of females that would be
born, on the average, to each of a group of women, beginning life together
and assuming that:

26
 • None of the dies before reaching the age in the upper limit of age group
• All of them experience, throughout their reproductive span the current
level of fertility
However this information may not always be available but the total num-
ber of births be known. In such a case an approximate value of GRR may be
computed from the following expression which is based on the assumption
that sex ratio at birth is more or less constant over all ages of mothers:
= T F R ∗ ( T otalT otal
number of f emale births
number of births
)
Gross reproduction rate may be used for comparing the fertility situation
in different regions. However, because of under registration of births and
wrong statement about the age of mother at the time of registration, the
value of gross reproduction rate may not be accurate. Also it assumes that
none of the newly born babies dies before the age , which inflates the number
of future mothers

Net reproduction rate (NRR)

NRR is another measure of population growth. Growth in population de-

pends on the balance between births and deaths. Thus net reproduction rate
is simply gross reproduction rate adjusted for the effects of mortality. It is
also a hypothetical figure giving the number of female babies that would be
born to a group of females beginning life together if they were subject to cur-
rent rates of fertility and mortality throughout their life span.As probability
is always < 1 therefore net reproduction rate < gross reproduction rate i.e.
gross reproduction rate serves as an upper limit to net reproduction rate.
If NRR = 1, it implies that a population of newly born girls will ex-

27
actly replace itself in the next generation provided the current fertility and
mortality rates prevail in future also. In this case population will have a
tendency to remain constant in size. If net reproduction rate is greater than
one, population has a tendency to increase as in this case each female will be
replaced by more than one daughter. Similarly if NRR is less than 1, pop-
ulation will tend to decline. However, net reproduction rate should not be
used for forecasting the size of future population as it assumes that current
rates of fertility and motility prevail in future also, an assumption, which is
not true. It also ignores the factor of migration and considers the age and
sex distribution of a hypothetical life table population that may be quite
different from that of the actual population.

2.1.4 Ratios used in demography

The most common ratios are the sex ratio and the dependency ratio.

Sex ratio at birth

Sex ratio relates the number of males to females in the same populations and
therefore measures the numerical balance between the sexes.

T otal males
( ) ∗ 100
T otal f emales

A sex ratio with a value more than 100 indicates an excess of male over the
female births. A value less than 100 indicates an excess of female over male.
Example 1 The total number of male births registered in Kenya in 1960 was
16218 and the number of female births was 14219. Find the sex ratio at

28
birth.
solution
16218
( ) ∗ 100
14219
= 114.06
114 male births per 100 female births in the population.
Example 2 The total number of births in kenya in 1970 was 24216 . Find
the total of females and males using the sex ratio at birth in the previous
example.
solution
114
N umberof malebirths = ( ) ∗ 24216
214
= 12900
100
N umberof f emalebirths = ( ) ∗ 24216
214
= 11316
Example 3 The sex ratio at birth in a given population is given as 113. If the
sex ratio at birth of 105 male births to 100 female births. Find the number
of female births per 100 male births when sex ratio at birth is taken as 105.

105 113
=( = )
100 x

100
X=( ) ∗ 113
105
= 107.6 =108
The calculation suggests that there is a relative under registration of about
8 percent of females in the registration system.

29
General sex ratio

This is the ratio of males to females in a given population. This ratio can be
obtained from census enumerations of the total population or from sample
surveys.
All males
GSR = ( ) ∗ 100
All f emales
Example 1 The number of males from 1989 census was 11516219 and the
number of females was 13217430. Find the general sex ratio.

All males
GSR = ( ) ∗ 100
All f emales

11516219
GSR = ( ) ∗ 100
13217430
= 87

Dependency ratio

It is defined as the ratio of youths under 15 years of age plus peresons aged
65 and above to adults aged 15 to 64 years. It indicates the relative predom-
inance of persons in the dependent ages in relation to those in the productive
ages as broadly defined in most social and economic systems.

(Children under 15 years) + (persons aged 65+)

dependency ratio = ( )∗100
persons aged 15 − 64 years

Example 1 In a census the population agd 0-14 as enumerated was 4414000

, those aged 65+ was 421231 and the population 15-64 years was 4835231 .

30
Find the deoendency ratio

(4414000) + (421231)
dependency ratio = ( ) ∗ 100
48335231

= 100
Note: There exists many other ratios in demographic analysis such as:-
ratio of birth to death, sex ratio at death , ratio of married females to married
males.

31
3 ERRORS IN DEMOGRAPHIC DATA
The accuracy of demographic statistics varies from one country to another.
The deficiencies are most in the developing countries because among other
problems of lack of administrative machinery, individual ignorance about
certain personal details and sometimes open hostility to some types of inquiry,
due to ignorance. Errors in the demographic data are mainly of two types,
namely, coverage and content errors.

• Coverage Errors – are due to persons being missed or counted more

than once. They are of two types. Individual of a given age may have
been missed by the census officials or erroneously included in it (i.e.
counted twice). The first type is called under enumeration at this age,
while the second type represents over enumeration, the balance of the
two types of errors represents net under enumeration at this age.

• Content Errors - refers to instances where the characteristics of a person

counted in a census enumeration or in the registration of a vital event
are incorrectly reported or tabulated. The five principal sources of
errors of content are:

1. The respondent - when he/she intentionally or unintentionally fail

to give the appropriate information required. A classic example is
misstatement of age or the case where widows or divorces report
themselves as single.

2. The enumerator - where he/she classifies a person incorrectly even

though give the correct information.

32
 3. The coding process - errors caused from the failure to allocate
information on the census or survey schedule to the proper code.

4. The editing process.

5. Errors may arise at all stage of the compiling process or during

tabulations.

3.1 ERRORS IN AGE DATA

Demographic data are usually classified by age and sex. Errors in age report-
ing are more frequent than errors in sex reporting. With reference to Kenya
in census/survey reporting of age, there are five major forms irregularities -

• Under reporting of children aged less than one year.

• A tendency to give an exact age of some legal significance e.g. voting

at election or marriage.

• Distinct over statement of age at very advanced ages.

• The reporting of some individuals as being of an unknown age.

• Age heaping.

These irregularities must be detected, adjusted or corrected before demo-

graphic data could be used for any meaningful analysis. Hence, the following
methods of evaluating demographic data or of detecting age irregularities.

33
3.1.1 Whipple’s Index of Digit Preference

This give the relative preference for digit ‘O’ and ‘5’ while reporting age in
the interval 23 to 62 years. It is computed as;

P 25 + P 30 + P 35 + P 40 + P 45 + P 50 + P 55 + P 60
WI = ∑ ∗ 100
1/5 PX

Where Px = the number of persons reporting their age as x years. If there

is no heaping at ‘0’ and ‘5’ the index will have a value of 100. If there is
complete heaping, the index will have a value of 500.
To test for heaping at ages 30, 40, 50, and 60 i.e. ages ending in ‘0’ the
index is computed as;

P 30 + P 40 + P 50 + P 60
WI = ∑ ∗ 100
1/10 PX

To test for heaping at ages ending in ‘5’ i.e. 25, 35, 45, and 55, the index
is computed as;

P 25 + P 35 + P 45 + P 55
WI = ∑ ∗ 100
1/10 PX

Limitations of Whipple’s index

• It does not measure preference for digits other than ‘0’ and ‘5’

• It considers only the arbitrary interval 23 to 62 years and not the entire
life span of 0 to 80 or 100 years.

• It does not take into consideration the decreasing nature of the age
distribution due to depletion by death.

34
 • It’s applicable only to single year’s data.

3.1.2 Myer’s Blended Index of Digit Preference

This index is used for evaluating single - year age – sex data. It can give the
extent of digit preference for all the digits 0, 1, 2, 3... 9. It can be used to
report errors for all ages 10 – 99 years.
Assumption
The underlying assumption of the method is that in the absence of sys-
tematic irregularities in the reporting of age, the blended sum at each termi-
nal digit should be approximately equal to 10 percent of the total blended
population. If the sum at any given digit exceeds 10percent of the total
blended population, it indicates over selection of ages ending in that digit
(i.e. digit preference). On the other hand, a negative deviation or sum that
is less than 10 percent of the total blended population indicates an under se-
lection of the ages ending in that digit (i.e. digit avoidances). If age heaping
is non-existent, the index would be approximately zero.
Example
Use the Myer’s blended index to assess the quality of age data given
below-
Example Using age distribution Use the Myer’s blended index to assess
the quality of age data given below-
Procedure for computations

1. Sum all the populations ending in each digit over the whole range i.e.
10-99

2. Sum figures between ages 20-99.

35
 3. Multiply the sums in (1) by coefficients; 1, 2, 3,4,5,6,7,8,9 and 10.

4. Multiply the sums in (2) by coefficients from 9 descending to 0 i.e. 9

,8 ,7, 6 , 5 ,4 ,3 ,2 ,1 ,0.

5. Add the product of (3) and (4) , to obtain the blended sum

6. Add up the blended sum in (5).

7. Find the percent (

8. Take the deviations of each

This result indicates the extent of concentration or avoidance of a partic-

ular digit.
56309

Limitations of Myer’s Blended Index

1. It has no sound theoretical basis

2. It does not capture other forms of age bias

3. It is not suitable for grouped data.

3.1.3 United Nations Age Sex Accuracy Index

This index which was proposed by the United Nation is used for evaluation
of five-year age-sex data. The index is also referred to as Joint Score. It has
three components;
(a) Average sex ratio score

36
 This score is obtained by fist calculating the sex ratio at each age group.
Successive differences irrespective of sign are added and averaged. Age –
specific sex ratio = 5Pxm X 100 5PXf 5
Pxm =males aged x to x + 5 5
pxf = females aged x to x + 5
(b) Average male age ratio score (M)
For each age group for males, calculate the age ratios computed as Age
ratio = 5Px X 100 ½ (5Px – 5 + 5Px + 5) The deviations from unity
irrespective of sign are added and averaged (M).
(c) Average female age ratio score (F)
For each age group for females, the age ratios are calculated using the
same formulae as for males. The deviations from unity irrespective of sign
are added and averaged (F). The index is then computed as: UNAI = 3(S)
+ M + F.
The reported age-sex data for a given population is presumed to be ac-
curate if the age-sex accuracy index is between 0 and 19.9, inaccurate if the
index is between 20 and 39.9, and highly inaccurate if the index is above 40.
Example —

37
4 Life Table
A life table gives the mortality experience of a hypothetical group of people
(called the cohort) starting life together and experiencing same mortality
conditions as given by the observed age specific death rate throughout their
lifetime. It shows how this cohort gets depleted through deaths at each age
till finally nobody is alive. One can compute the probability that a person
of some given age will live for a specific number of years. It also enables to
compute the average longevity per person. The data for the construction of
life tables is taken from census and death registers. As a number of factors
cause differential mortality, life tables may be constructed on the basis of each
one of them such as religion, sex, occupation etc. The life table has been a
key tool of actuaries for some 200 years and is the basis for calculating life
expectancy.

4.1 Uses of life tables

Life tables are extensively used by demographers, public health programmers,

socio-economic decision-makers, insurance agencies, actuaries, Government
etc. The fields of life insurance and actuarial science revolve around life
tables. These tables were originally devised to meet the requirements of
insurance agencies.

• They enable in deciding the amount of premium payable under different

policies for various age groups.

• They help in the computation of net reproduction rate so as to study

population growth.

38
 • They are used by the government for determining retirement benefits
for its employees, for knowing the number of senior citizens, for know-
ing the future size of the population for estimating the school going
population etc.

• In case information on cause of death is also available then this can

be incorporated in a life-table to find the probability of death due to a
specific cause.

• Further, one could also study the impact of this cause on expectation
of life.

4.1.1 TYPES OF LIFE TABLE

4.1.2 Generation of life table

The cohort or generation life table is usually constructed for groups still
living and stops at the present age.

4.1.3 Conventional life table

The period or cross sectional or conventional life table express the mortal-
ity experience that a hypothetical cohort would have if it experienced the
mortality rates observed in a given time.

4.2 Basic assumptions in the construction of a life table

1. Start with a hypothetical group of infants say 1,00,000 all born at the
same time.

39
 2. The deaths are distributed uniformly over the age interval (x, x+1) (an
assumption which is not valid for early years of life, especially for age
0).

3. There is no change in the cohort but for deaths i.e. it is closed to

migration.

4. At all ages individuals die according to a predetermined fixed schedule.

4.3 Description of various columns of a life table

A life table consists of the following columns:

1. x Age on the last birthday where x takes non-negative integral values.

2. lx Number of persons living at any specified age x.

3. l0 Cohort or radix of life table. Usually set to 10n

4. dx Number of persons who attain age x and die before reaching the age
(x+1) = l( x) − l( x + 1)

5. qx The probability that a person of exact age x will die before reaching
dx
the age (x + 1) = lx

6. px The probability that a person of exact age x will survive till the age
lx+1
(x+1) = lx

7. mx Probability that a person in the age group x to (x+1) will die while
2qx
in this age group. = 2−qx

40
 8. µx Ratio of instantaneous rate of decrease in lx to the value of lx; =
− dlogl
dxl
x

9. Lx Number of years lived in the aggregate by the cohort of l0 persons

between ages x and (x+1); = 21 (lx − lx+1 )This approximation assumes
that deaths occur, on average, half way in the age interval x to x+1.
Such is satisfactory except at age 0 and the oldest age, where other
approximations are used.

10. Tx Total number of years lived by the cohort after attaining age x
= Lx + Lx+1 + .......

11. e0x Complete expectation of life at age x is the average number of years
lived after age x.

12. e00 Expectation of life at birth.

13. ex Average number of complete years lived by the cohort after age
x.=e0x - 21

NOTE: qx column is called the pivotal column of a life table. Once the
values of l0 and qx are known, the entire life table can be constructed. The
2mx
values of qx can be obtained from the relationship 2+m x
where the values of
dx
mx are calculated from the relationship mx = Lx
on the basis of census data
and death registration data.
Consider a large group, or ”cohort”, of males, for example, who were born
on the same day. If we could follow the cohort from birth until all members
died, we could record the number of individuals alive at each birthday – age
x, say – and the number dying during the following year. The ratio of these

41
is the probability of dying at age x, usually denoted by q(x). It turns out
that once the q(x)’s are all known the life table is completely determined.
In practice such ”cohort life tables” are rarely used, in part because indi-
viduals would have to be followed for up to 100 years, and the resulting life
table would reflect historical conditions that may no longer apply. Instead,
one generally works with a period, or current, life table. This summarizes
the mortality experience of persons of all ages in a short period, typically one
year or three years. More precisely, the death probabilities q(x) for every
age x are computed for that short period, often using census information
gathered at regular intervals (every ten years in the Kenya.). These q(x)’s
are then applied to a hypothetical cohort of 100,000 people over their life
span to produce a life table. An example is given below.

42
Table 1: Life table for U.S. Males. From: National Center for Health Statis-
tics (1997).

x l(x) d(x) q(x) m(x) L(x) T(x) e(x)

0 100000 1039 0.0104 0.0104 99052 7182893 71.8
1 98961 77 0.0008 0.0008 98922 7083841 71.6
2 98883 53 0.0005 0.0005 98857 6984919 70.6
3 98830 41 0.0004 0.0004 98809 6886062 69.7
4 98789 34 0.0004 0.0004 98771 6787252 68.7
5 98754 30 0.0003 0.0003 98739 6688481 67.7
6 98723 27 0.0003 0.0003 98710 6589742 66.7
7 98696 25 0.0003 0.0003 98683 6491032 65.8
8 98670 22 0.0002 0.0002 98659 6392348 64.8
9 98647 19 0.0002 0.0002 98637 6293689 63.8
10 98628 16 0.0002 0.0002 98619 6195051 62.8
20 97855 151 0.0016 0.0016 97779 5211251 53.3
30 96166 197 0.0021 0.0021 96068 4240855 44.1
40 93762 295 0.0032 0.0032 93614 3290379 35.1
50 89867 566 0.0063 0.0063 89584 2370098 26.4
51 89301 615 0.0069 0.0069 88993 2280513 25.5
60 81381 1294 0.0159 0.0160 80733 1508080 18.5
70 64109 2312 0.0361 0.0367 62953 772498 12.0
75 51387 2822 0.0549 0.0565 49976 482656 9.4
76 48565 2886 0.0594 0.0613 47121 432679 8.9
80 36750 3044 0.0828 0.0865 35228 261838 7.1
90 9878 1823 0.1846 0.2041 8966 38380 3.9
43
100 528 177 0.3351 0.4080 439 1190 2.3
The columns of the table, from left to right, are:

• x: age

• l(x), ”the survivorship function”: the number of persons alive at agex.

For example of the original 100,000 U.S. males in the hypothetical
cohort, l(50) = 89,867 (or 89.867%) live to age 50. l(0) is known as the
radix of the life table usually set as10n

• d(x): number of deaths in the interval (x,x+1) for persons alive at age
x. Thus of the l(50)=89,867 persons alive at age 50, d(50) = 566 died
prior to age 51.

• q(x): probability of dying at age x. Also known as the (age-specific)

risk of death. Note that q(x) = d(x)/l(x), so, for example, q(50) = 566
/ 89,867 = 0.00630.

• m(x): the age-specific mortality rate. Computed as the number of

deaths at age x divided by the number of person-years at risk at age x.
Note that the mortality rate, m(x), and the probability of death, q(x),
are not identical. For a one year interval they will be close in value,
but m(x) will always be larger.

• L(x): total number of person-years lived by the cohort from age x to

x+1. This is the sum of the years lived by the l(x+1) persons who
survive the interval, and the d(x) persons who die during the interval.
The former contribute exactly 1 year each, while the latter contribute,
on average, approximately half a year. [At age 0 and at the oldest age,

44
 other methods are used; for details see the National Center for Health
Statistics (1997) or Schoen (1988). Note: m(x) = d(x)/L(x).]

• T(x): total number of person-years lived by the cohort from age x until
all members of the cohort have died. This is the sum of numbers in
the L(x) column from age x to the last row in the table.

• e(x): the (remaining) life expectancy of persons alive at age x, com-

puted as e(x) = T(x)/l(x). For example, at age 50, the life expectancy
is e(50) = T(50)/l(50) = 2,370,099/89,867 = 26.4.

Notes

1. Life expectancy is not the same as median survival time, the latter be-
ing the time at which only 50% of a cohort are still alive. For example,
of the 100,000 persons alive at age 0, 51,387 are alive at age 75, and
48,565 are alive at age 76. The median survival time at birth (age 0)
is thus between 75 and 76 additional years (and can be shown to be
75.5), while the life expectancy at birth is e(0) = 71.8 additional years.

2. The calculation of life expectancy for a person should not be confused

with predicting their survival time. While newborn U.S males have a
life expectancy of 71.8 years, any given U.S. male may die tomorrow
or live to age 100. One need not predict actual survival times in order
to compute life expectancy (the average survival time).

There are two major types of life table which can be constructed: Com-
plete life table A complete life table is a table in which the mortality expe-
rience is considered in single years of age throughout the life span. All the

45
columns of the life table are given for single years and it is extremely de-
tailed. Abridged life table An abridge life table is one in which the measures
are given not for every single year of age, but for age groups.

Example

to be done in class

4.3.1 Life tables for persons with risk factors

To construct a life table for a group of individuals with a common risk factor
(or factors), one may either

1. compute age-specific mortality rates from a suitable database, or

2. adjust the mortality rates in the basic life table by adding excess death
rates (EDRs) or multiplying by relative risks (RRs), either of which
may be available from published studies.

5 Graphical presentation

5.1 Population Pyramid

5.2 Lexis Diagram

Lexis diagrams provide a graphical representation of the relationships be-

tween demographic events in time and persons at risk and they also assist
in calculating rates � Every demographic event is characterized by two num-

46
bers: the time (e.g., year) at which it occurs and the age (or other duration
measure) of the person to whom it occurs

6 POPULATION PROJECTION

Introduction

Forecasts are the basis for all forms of planning for the future. Even a
complacent assumption that life will continue much as in the past is a form
of forecast. Demographic forecasts, in particular, are fundamental to any
form of social, economic or business planning. Suppliers of any good or
service, from maternity hospitals to the manufacturers of coffins, can only
plan ahead if they have some idea how many potential users of their services
they can expect to be located where.
Some phenomena can be forecast with perfect accuracy. For example,
we know exactly when and where future solar eclipses will occur. Other
phenomena, such as the weather, are inherently impossible to forecast with
any certainty for more than a few days into the future. It seems unlikely that
we will ever to be able to forecast future demographic developments with the
certainty that we can eclipses; we certainly cannot yet. Nevertheless, human
populations have two fundamental characteristics that reduce uncertainty
about how they will develop in the future:

• First, a substantial overlap exists between the current population and

the future population. For example, everyone who will be aged 25 or
more years in 25 years time has already been born.

47
 • Second, one fundamental aspect of the human condition is that every
year that passes we all get exactly one year older until we eventually
die.

These two facts constrain possible future developments in a population in

ways that often have no equivalent in other fields, such as economic forecast-
ing. Standard methods for making population projections take full advantage
of both facts.
This session provides an introduction to population projections and fore-
casting. It briefly discusses projection methods based on extrapolating pop-
ulation growth before describing the cohort component projection method in
more detail.

Population projection VS population forecast

A population projection can be defined as: A computational procedure to

calculate population size and structure at one time from population size and
structure at another, together with a specification of how change takes place
during the interim period.forecast of future population can be defined as a
projection based on assumptions that are predictive and considered to yield
the most probable estimates of the future development of a population. Thus,
while all forecasts of the population are projections, not all projections are
forecasts.
Population forecasts can also be contrasted with population estimates,
which are statistics on the current or past size and structure of a population.
Strategies for projecting populations
Two contrasting approaches exist for carrying out projections:

48
 Total methods
Total methods calculate trends in the size of the population as a whole
using a mathematical model of population growth. They may then distribute
this total into sub-groups in ratio to the current structure of the population
or an extrapolated forecast of its structure. Therefore, such approaches are
sometimes known instead as ratio methods of projection.
Cohort component methods
Cohort component methods project each age group, sex and other cat-
egory of interest separately. They then aggregate the results to obtain the
total population. The term cohort emphasizes that an age group is made up
of people born at the same time who go through life together. The size of a
cohort at one age (and date) is strongly predictive of its size at other ages
(and dates).
Many population projections combine both approaches, although projec-
tions dominated largely by the cohort component approach are by far the
most common. Nevertheless, cohort component methods require many more
input data and assumptions than total methods and may be inappropriate:

• if only estimates of total numbers are needed

• if the information required as input to component methods is lacking

• for very long-term projections.

Total methods of projection

Total methods of projection involve fitting a mathematical model to data

on past trends in the size of the population and using the fitted model to

49
extrapolate the population forward (or on occasion backward into the more
distant past).
The main steps involved in the procedure are to:

• select an appropriate model of the growth process

• estimate the parameters of the model from past estimates of the pop-
ulation

• extrapolate the fitted curves and read off the projected population.

The following are several mathematical functions that can be used to model
population growth

• Zero population growth

• arithmetic growth

• exponential growth

• logistic growth

• Geometric model

Zero population growth

zero population growth graph

The simplest model of population growth is the zero population growth
model, which assumes that the size of the population is unchanging. This
assumption implies that, even if one only has a single existing estimate of
the size of the population, one can project its size at other dates.

50
Arithmetic growth

The next simplest model is that of arithmetic or linear growth. This model
assumes that a constant numeric change occurs in the size of the population
in every period of the same length.A minimum of two estimates of the popu-
lation for different dates are needed to estimate the annual increment in the
population and project its size at other dates. The model can be fitted to
a longer series of estimates of the population by a means of a simple linear
regression of population size on time.
Thus, if P(t) refers to the population at time t and P(t+n) refers to the
population n years later:

P(t+n) = P(t) + a × n

where a is the constant annual increase in the population

(P(t+n) − P (t ))
a=
n

Exercise

If the population is 62,000 at time 275 and was 50,000 at time 225, what will
be the population at time 350?

Exponential growth

Instead of assuming that the population is growing by a constant amount,

the exponential model assumes that the population is growing at a constant
rate. This may be appropriate for expanding communities unaffected by any

51