See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327286019

The Models of Age-specific Mortality Rates and Their Patterns from Female Total
Population Counts
Article · August 2018

CITATIONS READS
0 573

1 author:

Nirmal Gautam
University of Southern Queensland
11 PUBLICATIONS 45 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Population Research View project

All content following this page was uploaded by Nirmal Gautam on 29 August 2018.

The user has requested enhancement of the downloaded file.

Journal of Population and Social Studies, Volume 26 Number 3, July 2018: 235 - 247
DOI: 10.25133/JPSSv26n3.017

The Models of Age-specific Mortality Rates and Their

Patterns from Female Total Population Counts

Nirmal Gautam1, Attachai Ueranantasun2 and Apiradee Lim3

Abstract
If net migration is negligible, population change in a country is governed by an increase due to births
and a decrease due to deaths. This study focuses on the mortality of female population based on their
total fertility rates. It develops a simple method to analyze population data with the aim to estimate
age-specific mortality models for 54 countries in the world and identify a similar pattern among
them. Specifically, the method involves estimating the mortality rates of those countries with
negligible net age-specific migration to form the common models. These models are then applied to
other countries by matching mortality at older ages to verify similar patterns. The results show three
different mortality models which imply the differences in the mortality patterns between the countries
studied.

Keywords
Age-specific female mortality; mortality patterns; population data

Introduction
Mortality pattern refers to human survival and longevity in any population. It is a key
indicator of health and development of the nation (Sharrow, Clark, & Raftery, 2014).
Changes in the population structure of a country have been historically regulated by
mortality (Schröder, Leeuwen, & Cameron, 2014). Therefore, reducing mortality has been a
major goal of population health policies with considerable investments in public health and
medical technology to extend longevity and improve public health (Bangha, 2013). It has
been documented that declines in mortality compensate for decreased fertility rates, so
mortality and fertility rates impact population growth considerably (Shelton, 2014). The
declining trend is particularly evident in child and maternal mortality rates in the studied
countries that have the appropriate records since the 1960s (Liu et al., 2015). Furthermore,
female mortality remains high, mostly in Sub-Saharan Africa and Southeast Asia, because of
poor health-related behaviors and socioeconomic status (Johri, Ridde, Heinmüller, &
Haddad, 2014; World Health Organization, 2015a).
While human mortality affects everyone and differs by country because of the discrepancies
in economic development, resulting from the industrial revolution and sanitary progress
(Meslé & Vallin, 2010). Additionally, mortality patterns of the countries heavily depend on
the mixture of other socioeconomic factors, such as incomes and educational levels. These

1 Department of Mathematics and Computer Science; Faculty of Science and Technology, Prince of
Songkla University, Pattani Campus
2 Department of Mathematics and Computer Science; Faculty of Science and Technology, Prince of
Songkla University, Pattani Campus Email: [email protected]
3 Department of Mathematics and Computer Science; Faculty of Science and Technology, Prince of
Songkla University, Pattani Campus

235
The Models of Age-specific Mortality Rates

differences in mortality rates are confirmed by a study which assesses social-economic

factors among countries in Western Europe to reveal their effects on morbidity and mortality
(Mackenbach et al., 1997). The differences in mortality are associated with inequalities in life
expectancy at birth by country, ranging from low life expectancy in low-income countries to
high life expectancy in high-income countries (World Health Organization, 2004).
Throughout the world, females have a longer life span and more life advantages than males
(Chaudhuri, 2015; United Nations, 2015a). The difference in the longevity and mortality rates
of females and males across the different countries is regulated by their biological,
behavioral and environmental factors (Stevens, Mathers, & Beard, 2013; Pirkis et al., 2016).
Globally, the total female population was 3.6 billion in 2015 (United Nations, 2015a), and
more than half of them are living in poverty and with poor health and nutritional status
(World Bank, 2012). In 2015, around 14,545 children and 822 mothers were dying daily due
to the unavailability of health care services, poor nutrition, maternal and child health
complication, violence, sexual abuse and other causes. More than 90% of them were from
low- and middle-income countries (World Health Organization, 2015b; United Nations
Children's Fund, 2016). In contrast, only 10% of those children and just only 1% of those
mothers were dying daily due to the behavioral, environmental and maternal complications
in high-income countries (World Health Organization, 2015a).
Females in general have negative health effects in their entire course of life due to lower
access to maternal and child health care. Hence, gaining knowledge of female mortality can
be beneficial in improving health policy, and thus improving the survival and extending the
longevity of females. Additionally, to estimate female mortality rate in the easiest way for an
international scale is difficult, because it requires a good record of registration system.
Moreover, female mortality plays an important role in the demographic analysis, which
helps to estimate and project the population size (Upadhyay et al., 2014). The results of the
analysis can contribute to better handling of population growth and managing aging
populations. One important characteristic of mortality is its age-specific rate (Brown, 2003).
Many researchers and international organizations have estimated the age-specific mortality
rate using different models and techniques from censuses, vital registrations, data and
household surveys (Clark & Sharrow, 2011; Wilmoth et al., 2012). Nevertheless, it has been
found that most of the developing countries and even some developed countries were
lacking complete vital registration data, and some demographic data required to calculate
age-specific mortality (Luy, 2012; Adegboye & Danelle, 2014). To handle this issue, the
census-based method from the stable population (Preston & Bennett, 1983), forward
projection technique and modified growth balance equation have been used for estimating
adult mortality (Preston, Coale, Trussell, & Weinstein, 1980; Bennett & Horiuchi, 1981).
These studies imply that age-specific mortality rate can be determined using only population
data from the census without death registrations. Even though the census-based method can
be implemented for determining the mortality rate, more than half of all countries in the
world do not have their own censuses with high quality acceptable data (Raftery, Chunn,
Gerland, & Ševčíková, 2013; US Census Bureau, 2016).
Based on these observations, the estimation of common models for age-specific mortality
rates for females across a group of countries can help develop understanding of the
mortality patterns for each of the studied countries. However, it is not easy to gather
information regarding complete vital registrations and censuses from each country. There
have been two major projects on gathering population and mortality data. Human Mortality
Database (2017) provides detailed information on age-specific deaths and mortality rates
among other population-related data. However, data for age-specific mortality rates from
these data sets are only available for 39 countries and most of them are high-income

236
 N. Gautam, A. Ueranantasun & A. Lim

countries. The other project is International Data Base from U.S. Census Bureau (US Census
Bureau, 2016). Even though the population data from this project are based on a projection,
not a census, the project provides accurate and reliable population data including fertility
rates and mortality indicators by gender for most countries around the world. Therefore,
data from International Data Base is useful for studying a large group of countries. Despite
the usefulness of other population data in this project, the information on age-specific
mortality is still lacking. Therefore, this study is intended to develop models for female age-
specific mortality for selected countries around the world using a simple and
straightforward technique using data from the US Census Bureau.

Methods
The study used data on female population in 54 countries, from 1990 to 2015, retrieved from
the US Census Bureau (2016). Only the world’s most populous countries were selected here,
and these 54 countries were chosen to equally represent Africa, Asia, and Western countries,
with 18 countries for each region. These regions were grouped based on birth, death, and
migration patterns. For example, Sub-Sharan countries showed similar patterns of birth and
death rates (Tabutin, Schoumaker, & Rabenoro, 2004) while South Asia tended to have
different patterns (Véron, Horko, Kneipp, & Rogers, 2008).
Previously, Coale, and Demeny (1966) used age-specific population data to develop their
classical model life tables for 192 countries, which were basically grouped according to
geographical regions. The countries within one geographical region usually shared the same
life table model. Nevertheless, there could be exceptions where a country from another
region is included in the life table model based on its region: for example, Taiwan and Japan
are in the West model (Coale & Demeny, 1966). Therefore, the concept of grouping
population trends based on their regions and similarities was adopted in this study. This
study was aimed at estimating the models for the mortality rates of each group of countries.
However, the mortality rates were not calculated directly, but rather extracted from the
population change. For this study, the population change was a combination of mainly
mortality and migration. First, population change was calculated using a simple method
based on the population data of a single calendar year and age of the females. In order to
estimate the age-specific population change for a female population, the following equation
was used:
N ( x  1, t 1)
Population change = 1
N(x,t)
Where, the function N(x, t) is the number of persons aged x in year t and N(x+1, t+1) is the
number of persons age x+1 in year t+1.
Subsequently, the estimated mortality rates for each country were calculated by minimizing
the values for all years. It was found that including data from all years from 1990 to 2015
gave the results that required smoothing, and the better results were obtained by selecting
fewer years. Therefore, only data for 1991, 1992, 1995, 1996, 2000, 2001, 2005, 2006 and 2008-
2015 was consequently selected. The calculated mortality rates were then plotted using a
cubic root scale to ensure the variance homogeneity assumption plausible. Single ages from
0 to 85 years were used and the mortality rate for all countries were grouped by patterns,
based on the level of fluctuations in the plots. In order to create a baseline, countries with the
least fluctuating patterns were selected and then divided into sub-groups based on similarity
of mortality patterns.

237
The Models of Age-specific Mortality Rates

For countries with only slightly deviating patterns, it was reasonable to assume that
migration was negligible. Net migration rates per 100,000 population and life expectance in
year 2010-2015 in each country are shown in Table 1 (United Nation, 2017). In each sub-
group, taking the median of population change provided a filtering mechanism to diminish
migration, and it therefore represented the baseline for population change in the sub-group.
This baseline is considered the model of mortality rate for each sub-group because of the
removal of migration from population change. These medians or the model of mortality
rates were then used in relation to population change from other levels of fluctuation to
form sub-groups in each level. A country with population change at old ages matching the
pattern of any particular model of mortality rate was then added into a corresponding sub-
group as it was reasonable to assume that migration at older ages was negligible. In each
sub-group, the difference between the population change and the model of mortality rate
indicated migration, either immigration or emigration for each age level. When the
population change for all 54 countries were categorized into groups according to the
medians or the mortality rates, it could be observed that the patterns of mortality rates could
systematically identify the model of mortality rates for each group. Countries with positive
population changes and the fluctuation of population change between current age group
and next age group approximate 0-200 per 100,000 female population are placed in smooth
pattern. Countries with negative population change and the fluctuation of population
change between 200-500 per 100,000 female population were classified into mild fluctuate
pattern while countries with negative population change and the fluctuation of population
change more than 500 per 100,000 female population were classified into highly fluctuated
pattern.

Table 1: Net migration rate and life expectancy for each country
Net migration rate (per 1,000) Averag
1990-1995

1995-2000

2000-2005

2005-2010

2010-2015

Life
e

Countries Migration expectancy rate

Smooth pattern 51.9

Nigeria Low -0.2 -0.2 -0.3 -0.4 -0.4 -0.3 63.6
Sudan Moderate 5.9 -4.9 -3.8 -6.6 -3.2 -2.5 62.6
Congo Moderate 1.7 3.3 2.9 4.7 -2.6 2.0 69.8
Tanzania Low 4.3 -1.3 -1.6 -1.2 -0.8 -0.1 60.2
Angola Low 2.2 -1.7 1.9 0.8 0.7 0.8 56.4
Cameroon Low -0.8 -0.7 -0.7 -0.6 -0.3 -0.6 75.6
Viet Nam Low -1.1 -0.6 -1.6 -2.0 -0.4 -1.1 75.7
China Low -0.1 -0.1 -0.3 -0.4 -0.3 -0.2 82.2
India Low -0.1 -0.1 -0.4 -0.5 -0.4 -0.3 68.6
Philippines Moderate -1.5 -2.1 -2.7 -3.3 -1.3 -2.2 82.5
Sri Lanka High -2.9 -5.0 -4.7 -5.2 -4.7 -4.5 66.0
Myanmar Moderate -3.3 -2.4 -5.3 -6.0 -1.9 -3.8 74.7
Brazil Low 0.0 0.0 0.0 0.0 0.0 0.0 73.8
Colombia Low -1.3 -1.0 -0.8 -0.6 -0.6 -0.9 77.0
Poland Low -0.8 -0.4 -1.0 -0.9 -0.4 -0.7 76.5
Mexico Moderate -2.9 -4.8 -5.5 -0.4 -0.5 -2.8 74.9
Romania High -4.5 -5.4 -4.3 -7.4 -3.0 -4.9 71.4
Venezuela Low 0.0 0.0 -0.1 -0.2 -0.5 -0.2
Mild fluctuating pattern 58.6
Uganda Low 1.3 -0.4 0.0 -0.9 -0.8 -0.2 58.7
Burkina Faso Moderate -3.2 -2.5 -2.0 -1.7 -1.5 -2.2

238
 N. Gautam, A. Ueranantasun & A. Lim

Net migration rate (per 1,000)

1995-2000

2000-2005

2005-2010

2010-2015
1990-1995
Life

Average
Countries Migration
expectancy rate

Ivory coast - - - - - 82.4

Nepal High 0.8 -4.1 -6.5 -7.8 -2.7 -4.1 61.8
Peru Moderate -2.6 -4.3 -4.8 -3.4 -1.6 -3.3 74.2
Ghana Low -0.2 -1.3 1.6 1.6 -0.4 0.3 61.7
Iran Low -7.2 1.9 -0.2 -1.5 -1.0 -1.6 68.6
Uzbekistan Low -3.0 -2.0 -1.9 -1.0 -0.5 -1.7 77.0
Indonesia Low -0.4 -0.3 -0.8 -0.6 -0.7 -0.6 68.6
North Korea Low 0.0 -0.1 -0.2 -0.2 -0.2 -0.1 51.9
Algeria Moderate -0.9 -1.1 -1.3 -2.1 -0.8 -1.2 75.3
Thailand Low -2.1 2.3 1.2 0.2 0.5 0.4 74.6
Egypt Low -1.5 -0.6 -0.2 -0.7 -0.6 -0.7 70.8
Russian Moderate 74.9
3.4 3.1 2.4 3.0 1.4 2.7
UK Moderate 81.0
0.7 1.7 3.3 6.6 3.1 3.1
Germany Moderate 80.5
6.6 1.7 2.0 0.1 4.4 3.0
Argentina Low 76.0
0.2 -0.4 -0.5 -0.4 0.1 -0.2
France Low 81.9
1.1 1.3 1.6 1.7 1.1 1.4
Highly fluctuating pattern
Madagascar Low 64.5
Ethiopia Low -0.2 -0.1 -0.1 -0.1 -0.1 -0.1 63.7
South Africa Moderate 5.6 -0.5 -0.2 -0.1 -0.1 0.9 59.5
Mozambique Low 3.7 1.4 2.5 2.5 3.0 2.6 56.1
Kenya Low 9.0 -0.6 0.3 -0.4 -0.2 1.6 65.4
Pakistan Moderate 1.8 -0.2 0.2 -1.0 -0.2 0.1 76.2
Morocco Moderate -1.7 -1.1 -0.9 -1.7 -1.3 -1.3 74.9
Bangladesh Moderate -3.6 -4.0 -4.4 -3.6 -1.8 -3.5 76.4
Malaysia High -1.4 -1.2 -2.2 -4.8 -3.2 -2.6 74.7
South Korea Low 3.1 4.5 5.3 5.3 5.3 4.7 59.5
Turkey Low 0.3 0.7 0.3 -0.6 0.7 0.3 74.8
Japan Low -0.4 -0.3 -0.1 -0.1 4.3 0.7 83.3
Taiwan Low 0.1 -0.2 0.3 0.4 0.6 0.2 69.8
USA Moderate -1.3 -2.6 1.9 2.2 1.5 0.3 62.8
Spain High 3.5 6.3 3.6 3.3 2.9 3.9 82.5
Australia High 1.6 4.5 13.4 10.0 -2.5 5.4 82.3
Canada High 4.0 4.1 5.9 10.6 8.0 6.5 56.4
Italy Moderate 4.9 5.1 6.5 7.4 6.5 6.1 82.4
0.5 0.8 5.6 3.4 0.9 2.2

Findings
After calculating the mortality rate from population change for 54 countries and plotting the
graphs reflecting population change, it was found the population change pattern could be
categorized into three groups based on migration: smooth, the mild fluctuating and the
highly fluctuating patterns. The smooth pattern group describes the smoothest pattern of
female mortality rates with sharing the low net migration countries, while the other two
groups showed mild and high fluctuating with intermediate and high net migration. Thus,
the name of each pattern was assigned by the fluctuations in plots for each group.

239
The Models of Age-specific Mortality Rates

Figure 1: Good patterns of population change

For the smooth pattern, the countries in this group were further divided into three sub-
groups based on the patterns in the population change plots. Figure 1 shows the smooth
pattern of population change of females for ages 0 to 85 years for all three sub-groups, with
the population change shown using a logarithmic scale. The ages below 1 year, 1 to 15 years,
16-59 years and over 59 years were defined as an infant, adolescence or young age,
productive age and old age respectively. The first sub-group on the left panel of Figure 1
reveals the population change for six African countries, including Angola, Cameroon, Congo
(K), Nigeria, Sudan and Tanzania. In this sub-group, all the countries have similar pattern of
population change in infant and old age categories. However, in the productive age, the
population change shows wider dispersion among these countries. The graph in the middle
panel of Figure 1 shows Myanmar, China, India, the Philippines, Sri Lanka and Vietnam. In
this plot, the variation of population change in these six Asian countries mostly is in the
adolescence, productive and old ages. However, in the adolescence and productive ages, the
population change shows fluctuations among these countries. The right panel of Figure 1
shows Brazil, Columbia, Mexico, Poland, Romania and Venezuela, with the least change in
population at young and old ages. In this sub-group, Poland showed a highly negative
population change at around 10 years of ages.
The median of each sub-group was created and these medians were considered the models
of mortality rates: Africa model, Asia model and West model according to the common
location of the countries in each sub-group. The models of mortality rates for all three sub-
groups are shown in Figure 1 as dotted lines. The shape of the model in each group indicates
the common pattern of mortality rate for the countries of that group. In the Africa model, the
model started at a high level before dropping significantly from infant to the age of 10 years.
It then started to go up sharply until age 35 after which the slope was less steep. The slope
changed again at age 60 when the population change went up more rapidly again. In the
Asia model, the model of mortality rate started at a lower level than in the previous model

240
 N. Gautam, A. Ueranantasun & A. Lim

and dropped with a hook shape from infant to 20 years of age. The rate then had a gentle
slope with small oscillation from 20 to 40 years and it went up exponentially afterwards. The
model began as lowest among the three models before dipping until age 10. After that, it
moved up with an exponential rate towards the end. Additionally, there was little
fluctuation between ages 20 and 50 in this model.
Figure 2: Wobbly patterns of population change

The mild fluctuating patterns in Figure 2 represent 18 countries. The three models of
mortality rate (Africa, Asia and West models), from the smooth patterns, were used to
categorize these countries by matching them with the models by considering similar
patterns at old ages. The results showed that all of the 18 countries can be categorized by the
three models, thus forming three sub-groups. The left panel of Figure 2 represents the Africa
model assigned to the four countries, Burkina Faso, Ivory Coast, Nepal and Uganda. All
these countries had almost similar patterns of population change at the infant and old age,
while in the productive age of females the population change had moderately fluctuating
patterns. The middle panel shows the Asia model matches Ghana, Indonesia, Iran, North
Korea, Peru and Uzbekistan. In this sub-group, the estimated female population change
rates are often unsymmetrical, mostly in young and productive ages, but in the old age
group, the countries showed quite similar patterns. Only Iran and North Korea had a
negative change in population, mainly around ages 15 to 40 and from 40 to 60 respectively.
The right panel of figure 2 consists of eight countries matching West model at old ages,
namely Algeria, Argentina, Egypt, France, Germany, Russia, United Kingdom and Thailand.

241
The Models of Age-specific Mortality Rates

In this sub-group, all the countries had a negative population change except for Egypt and
Thailand.

Figure 3: Bad patterns of population change

The highly fluctuating pattern cases shown in Figure 3 are also reflected in 18 countries, with
severe and erratic fluctuations patterns of population change. Similar to the previous
process, these countries were classified into sub-groups by matching with the three models.
In the left panel of Figure 3, the six countries, Ethiopia, Kenya, Madagascar, Mozambique,
Pakistan and South Africa, are in the same group matching the Africa model while the Asia
model is matched with five countries, namely Bangladesh, Malaysia, South Korea, Turkey
and Morocco. In the right panel of Figure 3, the West model is matched with seven countries:
Australia, Canada, Italy, Japan, Spain, Taiwan and the US. All these countries in the latter
group showed erratic fluctuations of population change with negative deviations.

Discussions
The present study provided estimates of the models of female mortality based on the latest
population data by calculating a population change. The population change of these 54
world’s most populous countries were divided into three categories: African model, Asian
model and Western model. It was shown based on their patterns, all the 54 countries could
be systematically classified into groups based on medians or the models. The patterns in the
models could be considered estimates of the mortality rate. Even across different countries,
humans share virtually similar patterns of mortality rates, higher at infant and at old age

242
 N. Gautam, A. Ueranantasun & A. Lim

while lower during adolescence and productive ages, in a normal situation (absence of wars
or natural catastrophes). The major differences in population change between countries are
mostly due to human movements or migration, and due special circumstances such as
outbreak of a disease. The use of medians as common models for mortality rates was
confirmed by the findings that indicated consistent results. It was shown that the countries
sharing the same characteristics as the model shared characteristics which distinguished
them from countries assigned to other models. In the Africa model, a higher mortality rate
was found among infants, adults and older females compared with Asian and Western
models. In the Asian model, the estimated mortality rate was clearly lower than in the Africa
model, but higher than in the Western one. These findings are consistent with Hill, You,
Inoue, and Oestergaard (2012), where Africa showed higher rates of the death for children,
while Asia/Pacific and Western countries had the lowest. In Africa, 1 out of 12 children dies
before reaching their fifth birthday, while it is 1 death among 19 children in Southeast Asia.
In developed countries, itis 1 out of 147 children, a considerably lower mortality rate (United
Nations Children's Fund, 2015). Other studies have found that countries in Africa and
Southeast Asia have the highest mortality rates for adult and old-aged females due to high
prevalence of HIV/AIDS, poverty, malnutrition, maternal complications, violence against
women, and power and decision-making in African and Asian countries (United Nations
Programme on HIV/AIDS, 2013; United Nations, 2015a). In contrast, the mortality levels
among females in western countries have been declining for all ages. This is due to
considerable improvements in public health policies and rapid economic development. The
same situation with respect to female mortality has been well documented in other studies,
where it is evident that females face a far worse situation than their male counterparts.
Hence, in the low and middle-income countries, females are at an elevated risk of dying in
their entire lives (Hill et al., 2007; Sawyer, 2012; Stevens, Mathers, & Beard, 2013 Alkema et
al., 2016). A joint study by the World Health Organization, (2015b), and United Nations
Children's Fund, (2015) revealed that females are dying due to lack of basic health care, poor
and inadequate education, poor nutrition and violence in the developing countries of Sub-
Saharan Africa and Southeast Asia, while another study also reported high female mortality
rates for the same regions. Therefore, it can be inferred that the mortality rate is dependent
on the country’s level of socio-economic and cultural development (Liang et al., 2010; Bayati
et al., 2016).
This study developed a model of female mortality rates based on patterns in the populations
of several countries. The patterns were classified into three categories: smooth, mild
fluctuating and highly fluctuating. The smooth patterns had fine patterns of population
structure, while the mild fluctuating patterns showed some deviations of patterns of
population change among the productive age females, and the highly fluctuating patterns
had severe fluctuations patterns throughout the ages from 0 to 65. These patterns also
helped in estimating mortalities, as explained earlier. If the population change is positive, it
is reasonable to assume that the net age-specific migration is negligible. In such a case, the
pattern of population change in the females directly reflects the estimated mortality rate.
This is because infants and older people do not migrate due to inherent bias and other
specific issues such as physical challenges and policies (Angel, 2003; World Health
Organization, 2016). Therefore, the changing population figures in these age groups are not
affected by migration but rather by death rate.
In the smooth pattern, all of the countries had a positive fluctuation pattern in the
population change, so it is reasonable to assume that migration is negligible. Therefore, these
countries provide the most reliable estimates of female population changes, interpreted as
mortality. In the mild fluctuating and the highly fluctuating patterns, most countries had
irregular negative fluctuation patterns of the population change. If the

243
The Models of Age-specific Mortality Rates

population change is negative, this reflects immigration. This mostly happens at younger
ages for females in high and some middle-income countries due to net positive migration
into these countries. Therefore, UK, USA, Canada, Australia, Germany, France, Italy, Russia,
Spain, Turkey, Taiwan, Algeria, Kenya, South Africa, Bangladesh, Iran, Malaysia, Pakistan
and South Korea can be considered hosts for immigrants of productive age from other
countries. This finding is consistent with prior literature. A United Nations report (2015b)
showed that more than 48% of females migrated to western countries, and some immigrants
went to African and Asian countries. This is because these countries have smooth income,
high literacy levels and good medical facilities. Therefore, these countries have hosted a
large number of immigrants, both legal and illegal ones, and refugees (United Nations,
2015a). In contrast, this type of migration stops at old ages among the female population.
Angel (2003) explained that this scenario is due to the fact that at higher ages, the
immigrants face greater distress having little savings and inadequate health insurance
coverage.
In low and most middle-income countries, fluctuations of pattern were found to be at low
level. It is again reasonable to assume these countries have negligible immigration. If the
immigration is negligible, then it is reasonable to accept the population change depends on
mortality rate and possibly emigration in some cases. Despite this, Morocco showed an exact
opposite fluctuation pattern with a repeatedly bumping pattern of population change at
productive age. This could be due to combined emigration and mortality. This is confirmed
by previous studies that during the period 1990-2015, a large number of emigrants from
Morocco fled to other countries as refugees or migrant labor and transiting to different
destinations. In addition, around 18.43 per 100 female adults also died due to maternal and
child health complications, and other reasons in Morocco (United Nations Economic
Commission for Africa, 2016; World Data Atlas, 2016).

Conclusion
This study showed female mortality patterns in 54 most populous different countries in the
world. This result, in combination with fertility data, is beneficial for exploring female
population dynamics for these countries. Furthermore, the countries in each model shared
the same pattern of mortality rate, and there could be a similar demographic pattern among
countries in the same region. Therefore, it is an indication the countries might also share a
common socioeconomic, health or development situations which can be explored in future
studies. With the exception of Japan and Taiwan in West model, it could be implied these
two countries possess similar socioeconomic, health or development situations with other
countries in their respective mortality pattern model.
Although the method used in this study was aimed at providing estimates of female
mortality rates for developing and developed countries based on their population change, it
is more suitable for the least developed and developing countries where migration is
relatively low. In high income countries, where immigration is fairly high except for old age,
this approach is not suitable to study population change among the young and the
productive female population. Another restriction of this study is that the accuracy of the
estimates can be disputed because they relied mainly on the accuracy of the source data.
While the US Census Bureau has provided a great deal of reliable projected population data,
there are many factors and turning points that could affect population projections, and in
turn, the estimated population change and mortality rates. Further studies including a large
sample size of countries might help capture model patterns and increase the understanding
of characteristic of population change and mortality rates.

244
 N. Gautam, A. Ueranantasun & A. Lim

In summary, in this study, age-specific mortality rates were estimated from the population
data by using a simple population change method. This method can be employed to reveal
the mortality rate if the countries experience low net migration. For the high migration
countries, the effects of migration must be filtered out by using the median or model for
more reliable cases before estimation of the mortality rates. The study found higher female
mortalities in African countries rather than in Asian and Western countries across all age
groups of the female population.

Acknowledgement
The authors express countless thanks to our Emeritus Prof. Don McNeil for his guidance,
encouragement, support and statistical advice. Also the authors would like to thank
Thailand's Education Hub for Southern Region of ASEAN Countries (TEH-AC 068/2015) for
providing funding, and acknowledge the US Census Bureau of Statistics for providing the
annual population data of females.

References
Adegboye, O. A. & Danelle, K. (2014). Causes and patterns of morbidity and mortality in Afghanistan:
Joint estimation of multiple causes in the neonatal period. Canadian Studies in Population, 41(1-
2), 165-179.
Alkema, L., Chou, D., Hogan, D., Zhang, S., Moller, A. B., Gemmill, A.,… & Say, L. (2016). Global,
regional, and national levels and trends in maternal mortality between 1990 and 2015, with
scenario-based projections to 2030: A systematic analysis by the UN maternal mortality
estimation inter-agency group. The Lancet, 387(10017), 462-474.
Angel, J. L. (2003). Devolution and the social welfare of elderly immigrants: Who will bear
the burden? Public Administration Review, 63(1), 79-89.
Bangha, M. W. (2013). Estimating adult mortality in Cameroon from census data on household
deaths: 1976-1987. African Population Studies, 23(2), 223-247.
Bayati, M., Vahedi, S., Esmaeilzadeh, F., Kavosi, Z., Jamali, Z., Rajabi, A. & Alimohamadi, Y. (2016).
Determinants of maternal mortality in Eastern Mediterranean region: A panel data analysis.
MedicalJournal of the Islamic Republic of Iran, 30, 360.
Brown, J. R. (2003). Redistribution and insurance: Mandatory annuitization with mortality
heterogeneity. Journal of Risk and Insurance, 70(1), 17-41.
Bennett, N. G. & Horiuchi, S. (1981). Estimating the completeness of death registration in a
closed population. Population Index, 47(2), 207-221.
Chaudhuri, S. (2015). Excess female infant mortality and the gender gap in infant care in Bihar, India.
Feminist Economics, 21(2), 131-161.
Clark, S. J. & Sharrow, D. J. (2011). Contemporary model life tables for developed countries. An
application of model-based clustering. In center for statistics and the social sciences:
University of Washington. 107, 1–38.
Coale & Demeny, P.G. (1966). Regional model life tables and stable populations. Princeton
university press, office of population research, USA.
Hill, K. & Pebley, A. R. (1989). Child mortality in the developing world. Population and
Development Review, 15(4), 657-687.
Hill, K., Thomas, K., AbouZahr, C., Walker, N., Say, L., Inoue, M. & Maternal Mortality Working
Group. (2007). Estimates of maternal mortality worldwide between 1990 and 2005: an
assessment of available data. The Lancet, 370(9595), 1311-1319.
Hill, K., You, D., Inoue, M. & Oestergaard, M. Z. (2012). Child mortality estimation: Accelerated
progress in reducing global child mortality, 1990–2010. PLoS Medicine, 9(8), e1001303.

245
The Models of Age-specific Mortality Rates

Johri, M., Ridde, V., Heinmüller, R. & Haddad, S. (2014). Estimation of maternal and child mortality
one year after user-fee elimination: An impact evaluation and modelling study in Burkina Faso.
Bulletin of the World Health Organization, 92(10), 706-715.
Liang, J., Zhu, J., Dai, L., Li, X., Li, M. & Wang, Y. (2010). Maternal mortality in China, 1996–2005.
International Journal of Gynecology & Obstetrics, 110(2), 93-96.
Liu, L., Oza, S., Hogan, D., Perin, J., Rudan, I., Lawn, J. E. & Black, R. E. (2015). Global, regional, and
national causes of child mortality in 2000–13, with projections to inform post-2015 priorities: An
updated systematic analysis. The Lancet, 385(9966), 430-440.
Luy, M. (2012). Estimating mortality differences in developed countries from survey information on
maternal and paternal orphanhood. Demography, 49(2), 607-627.
Mackenbach, J. P., Kunst, A. E., Cavelaars, A. E., Groenhof, F., Geurts, J. J. & EU Working group on
socioeconomic inequalities in health (1997). Socioeconomic inequalities in morbidity and
mortality in Western Europe. The Lancet, 349(9066), 1655-1659.
Meslé, F. & Vallin, J. (2010). Mortality patterns and their implications. Demography, I, 185.
Pirkis, J., Currier, D., Carlin, J., Degenhardt, L., Dharmage, S. C., Giles-Corti, B. & Keogh, L. (2016).
Cohort profile: Ten to men (the Australian longitudinal study on male health). International
Journal of Epidemiology, 46(3), 793-794i.
Preston, s. h. & Bennett, n. g. (1983). A census-based method for estimating adult mortality.
Population Studies, 37(1), 91-104.
Preston, S., Coale, A. J., Trussell, J. & Weinstein, M. (1980). Estimating the completeness of reporting of
adult deaths in populations that are approximately stable. Population Index, 46(2), 179-202.
Raftery, A. E., Chunn, J. L., Gerland, P. & Ševčíková, H. (2013). Bayesian probabilistic projections of
life expectancy for all countries. Demography, 50(3), 777-801.
Sawyer, C. C. (2012). Child mortality estimation: Estimating sex differences in childhood mortality
since the 1970s. PLoS Medicine, 9(8), e1001287.
Schröder, A., van Leeuwen, A. & Cameron, T. C. (2014). When less is more: Positive population-level
effects of mortality. Trends in Ecology & Evolution, 29(11), 614-624.
Sharrow, D. J., Clark, S. J. & Raftery, A. E. (2014). Modeling age-specific mortality for countries with
generalized HIV epidemics. PloS One, 9(5), e96447.
Shelton, J. D. (2014). Taking exception. Reduced mortality leads to population growth: An
inconvenient truth. Global Health: Science and Practice, 2(2), 135-138.
Stevens, G. A., Mathers, C. D. & Beard, J. R. (2013). Global mortality trends and patterns in older
women. Bulletin of the World Health Organization, 91(9), 630-639.
Tabutin, D., Schoumaker, B. & Rabenoro, M. (2004). The demography of Sub-Saharan Africa from the
1950s to the 2000s. Population, 59(3), 455-555.
United Nations Programme on HIV and AIDS (2013). Report on the global AIDS epidemic. Retrieved on
January 20, 2017 from http://www.unaids.org/sites/default/files/media_asset/UNAIDS_
Global_Report_2013_en_1.pdf.
United Nations. (2015a). World's women trends and statistics. Retrieved on March 12, 2017 from
https://unstats.un.org/unsd/gender/downloads/worlds women 2015 _ report.pdf.
United Nations. (2015b). Report of international migration. Retrieved on August 20, 2017 from
http://www.un.
org/en/development/desa/population/migration/publications/migrationreport/docs/Migr
ationReport2015_Highlights.pdf.
United Nations Children's Fund. (2015). Levels and trends in child mortality. Retrieved on March 20,
2017 from https:// www.unicef.org/publications/files/Child_Mortality_Report_2015_
Web_9_Sept_15.pdf.
United Nations Children's Fund. (2016). A fair chance for every child: State of the world's children.
Retrieved on February 20, 2017 from https://www.unicef.org/publications/files/UNICEF_
SOWC_2016.pdf.
United Nations Economic Commission for Africa. (2016). Demographic profile of African countries.
Retrieved on February 20, 2017 from
http://www.uneca.org/sites/default/files/PublicationFiles/demographic_profile_rev_
april_25.pdf.
United Nations. (2017). World population prospective: Population division. Retrieved on June3, 2018 from
https://esa.un.org/unpd/wpp/dataquery/.

246
 N. Gautam, A. Ueranantasun & A. Lim

Upadhyay, U. D., Gipson, J. D., Withers, M., Lewis, S., Ciaraldi, E. J., Fraser, A. & Prata, N. (2014).
Women's empowerment and fertility: a Review of the literature. Social Science & Medicine, 115,
111-120.
US Census Bureau. (2016). International data base. Retrieved on January 20, 2017 from https://www.
Census.gov/population/international/data/idb/information Gateway.php.
Véron, J., Horko, K., Kneipp, R. & Rogers, G. (2008). The demography of South Asia from the 1950s to
the 2000s: A summary of changes and a statistical assessment. Population, 63(1), 9-89.
World Bank. (2012). World development report: Gender equality and development. Retrieved on
June20, 2017 from
https://siteresources.worldbank.org/INTWDR2012/Resources/.../Complete-report.pdf.
Wilmoth, J., Zureick, S., Canudas-Romo, V., Inoue, M. & Sawyer, C. (2012). A flexible two-
dimensional mortality model for use in indirect estimation. Population Studies, 66(1), 1-28.
World Data Atlas. (2016). World and regional statistics. Retrieved on June 12, 2017 from
https://knoema.com/atlas.
World Health Organization. (2004). Report on changing history: Trends of HIV infections, therapy and AIDS.
Retrieved on July 25, 2017 from http://www.who.int/whr/2004/en/report04_en.pdf.
World Health Organization. (2015a). Under-five mortality: Global health observatory data. Retrieved on
July 20, 2017 from
http://www.who.int/gho/child_health/mortality/mortality_under_five/en/.
World Health Organization. (2015b). Trends in maternal mortality: Estimates from WHO, UNICEF,
UNFPA, World Bank Group and the United Nations Population Division. Retrieved on August 19,
2017 from http://www.afro.who.int/sites/default/files/2017-05/trends-in-maternal-
mortality-1990-to-2015.pdf.
World Health Organization. (2016). Migration and health key issue: Refugees and migrants: common health
problems. Retrieved on October 14, 2017 from http://www.euro.who.int/__data/assets/pdf_
file/0005/293270/Migration-Health-Key-Issues-.pdf?ua=1.

247

View publication stats