Vienna Yearbook of Population Research 2016 (Vol. 14), pp. 67–87
Population ageing dynamics in the North Atlantic
region of the Arctic
Anastasia Emelyanova and Arja Rautio∗
This paper contributes to our understanding of the demographic developments and
the transition to older age structures in the sparsely populated Arctic region: in
Iceland and in the two Danish autonomous territories of the Faroe Islands and
Greenland. We compare the population ageing dynamics of the region with those of
mainland Denmark for the 1980–2015 period. We also examine whether population
ageing has been developing diﬀerently in the communities of the North than in
Denmark, and shed light on the question of whether a regionally speciﬁc policy
approach to population ageing is required. In our study, ageing is measured by
applying a dual methodology. The two sets of indicators are based on calculations
of “chronological” and “prospective” ages. The latter is an innovative approach
developed by Sanderson and Scherbov (2008) that considers improvements in
life expectancy over time. Our results show that the size of the North Atlantic
region’s older population is well below the Danish national average. According
to chronological indicators, the ageing rates have been rising in recent years.
Prospective indicators, which take into account changes in population longevity,
also provide information about competing trends in population rejuvenation. In
addition, the prospective approach reveals a cross-territorial convergence in recent
decades, as well as a slower pace of ageing that can be accounted for in policy
∗Anastasia Emelyanova (corresponding author), International Institute for Applied Systems Analysis
(IIASA), Laxenburg, Austria
Arja Rautio, Centre for Arctic Medicine, Thule Institute, University of Oulu, Finland
68 Population ageing dynamics in the North Atlantic region of the Arctic
1.1 Study aims
The dynamics of “getting older” are central to “every single aspect of planning at
every single imaginable level: from the individual to the workplace to the societal
to the global levels” (Leeson 2002, 5). Indeed, the phenomenon of population
ageing, which is currently taking place almost everywhere in the world, is unfolding
on a scale hitherto unseen in human history. The question of how societies
should deal with population ageing has sparked heated debates across political and
scientiﬁc forums, especially in countries where life expectancy is very high, and the
population has reached its highest-ever point in terms of maturity (World Health
Organization 2011). At the same time, in the existing research, ageing that occurs at
a subnational and at smaller territorial levels has received considerably less attention
than ageing in major world regions and across countries.
Past research has shown that the rate of population ageing observed within remote
communities of the Arctic region diﬀers, sometimes signiﬁcantly, from the average
rates of the countries to which those communities belong, and from the rates of the
southern territories of Arctic countries. This pattern has been observed in various
parts of the Arctic, such as in the Barents Euro-Arctic region (Danilova et al. 2011;
Nummela et al. 2011; Emelyanova and Rautio 2012; Emelyanova and Rautio 2016),
the Russian North (Danilova and Golubeva 2011; Emelyanova and Rautio 2013),
the Nordic region (Nordregio publications, cited below), the North American Arctic
(Moore and Pacey 2004; Hamilton and Mitiguy 2009; Wilson et al. 2012; Lewis
2013), and across the Arctic as a whole (Larsen and Fondahl 2014; Emelyanova
2015). These authors have generally found that the ageing process in the sparsely
populated communities of the Arctic can transform those communities at every level,
and can profoundly distort the age composition of local populations. Subnational-
level data from these studies suggest that policy action on ageing that takes into
account the particular characteristics of the North is needed, and that modiﬁcations
of policy action at the national level may be required if cross-territorial diﬀerences
are found to be substantial.
The North Atlantic region of the Arctic includes Iceland and two autonomous
territories of Denmark: namely, the Faroe Islands and Greenland (Figure 1). These
areas have many common features. For example, all of these areas have (i) a
closed geographical location, (ii) a high degree of peripherality and isolation,
(iii) geopolitical status as part of Europe, (iv) an egalitarian social structure, (v) a
parliamentary system of government, (vi) a small domestic market with ﬁshing
being a predominant source of income, and (vii) a population living in either
relatively small urban centres or remote settlements. There is something of a gap
in the demographic literature on this part of the Arctic, especially with regard to the
ageing of its population. We aim to ﬁll this gap with this study.
Several papers published in English have contributed to our understanding of
the demography of coastal societies of the North Atlantic (Foss and Juvkam 2005;
Anastasia Emelyanova and Arja Rautio 69
The area of the study: North Atlantic, including Iceland, Greenland, and the Faroe
Source: The North Atlantic Marine Mammal Commission, http://www.nammco.no/.
Gløersen 2005; Rauhut et al. 2008; Hansen et al. 2011; Rasmussen 2011; Hansen
et al. 2012b; H¨
om and Roto 2013, Roto et al. 2014, Roto and Rasmussen
2016). However, relatively few of these papers have addressed the issue of societal
ageing, even though population ageing, along with urbanisation, are among the most
pronounced demographic trends of recent years.
According to the sources cited above, the demographics of the North Atlantic
region have changed dramatically over the past 40 years. Migration has been a
strong driver of this change, with many consequences for the age structures of small
communities. Currently, as in the past, large-scale industrial projects and recurrent
ﬁnancial crises in the region are triggering waves of migration. For instance, the
70 Population ageing dynamics in the North Atlantic region of the Arctic
main demographic crisis in the Faroe Islands is related to changes in the ﬁsheries
sector. While the sector was successful in the 1970s, it experienced a collapse
associated with overﬁshing in the 1990s, which in turn led to overinvestment,
unemployment, and out-migration. As a result of heavy emigration between 1989
and 1994, the population of the Faroe Islands declined by 10%. Since the 2000s,
positive economic changes have stabilised migration. In the case of Greenland, the
island moved from Danish colonial status to Home Rule in 1979. This new status led
to a large wave of out-migration, and the emigration rate of native Greenlanders did
not decrease until more attractive workplaces were established and investments in
education were made in the territory. Thousands of people emigrated from Iceland
in 2008–2009 in the wake of the ﬁnancial collapse of 2008; this was the largest wave
of migration from Iceland since 1887.
These historical events have caused the populations of the North Atlantic region
to age rapidly. Although the region still has higher birth rates than most European
countries, emigration rates have been especially high among people of reproductive
ages. The older population has risen most sharply in Iceland and in Greenland,
where the age structures had been young, but changed due to signiﬁcant outﬂows of
people of childbearing age, and particularly of young women pursuing educational
and career opportunities elsewhere. As these trends have distorted the birth-death
balance, societal ageing has become particularly intense in remote coastal areas of
the North Atlantic.
In light of these developments, the aim of this paper is to construct a
comprehensive cross-regional proﬁle of demographic ageing in the North Atlantic
for the 1980–2015 period using the available background data. We identify the
major demographic events that have been contributing to the shift towards the ageing
of the North Atlantic populations. In addition to using the traditional approach
to measuring ageing, we are rethinking ageing trends, and thus apply a new
“prospective” methodology. This approach allows us (i) to identify the oldest and
the youngest areas in the region, as well as the areas that are ageing the fastest and
the slowest (or that are, by contrast, rejuvenating); (ii) to compare the ageing rates
of the Danish autonomous territories to those of mainland Denmark; and, ﬁnally,
(iii) to examine whether the ageing trends are converging or diverging across the
countries and territories of the North Atlantic. The analysis will be of interest to
demographers, social scientists, and policy planners. Appendix 1 summarises the
ageing rates for the North Atlantic and Denmark for both sexes for the period under
In this study, we measure population ageing by applying a dual methodology, and
compare the results of the traditional and the new “prospective” approaches to
measuring ageing. The main diﬀerence between the two methodologies is the age
on which these measures are based: the chronological age or the prospective age.
Anastasia Emelyanova and Arja Rautio 71
The traditional indicators of population ageing include the median age (MA), the
share of people over the age of 60 (Prop 60+), the ageing index (AI), and the old-
age dependency ratio (OADR) (e.g., monitored in United Nations 2015). The MA
is represented by a line drawn to numerically divide the population into two equal
parts: one older and one younger than the age at the dividing line. The Prop 60+
relates the population aged 60 and older to the total population. The AI refers to the
percentage of people aged 60 and older divided by the number of children aged zero
to 14. The OADR relates the number of people aged 60 and older to the number of
people aged 15 to 59.
These indicators are based on chronological age; that is, on the number of years
already lived. As these indicators are monitored internationally and are cited in
most studies on societal ageing, they allow researchers to make broad regional
and even global comparisons of ageing trends. The computation techniques used
to construct these indicators are relatively simple, as they involve subdividing
population statistics by age and sex. In this text, we refer to results related to these
indices as having been measured in a “traditional” or a “chronological” way.
Traditional measures are based on a simpliﬁed understanding of ageing, with
chronological age being the only parameter considered. However, a number of
scholars have recently argued that these indicators are too limited (Denton and
Spenser 2000; Sanderson and Scherbov 2008, 2010). According to these authors,
the age of an individual is not commensurable across diﬀerent historical points
in time. For example, a Greenlander who was aged 60 in 1900 had far fewer
remaining years of life than his 60-year-old descendant living today due diﬀerences
in life expectancy (LE), health, and well-being during these two periods. These
observations have led scholars to rethink how we measure age and ageing, with
new metrics of ageing being introduced by Scherbov and Sanderson (2008). These
new measures incorporate changes in people’s characteristics beyond traditional
chronological age, such as prospective age and changes in remaining life expectancy
If we are considering purely chronological age, we know that a person who had
reached the age of, say, 50 in 1900 or in 2015 had lived for precisely 50 years. If,
however, we are considering prospective age, we cannot assume that the person’s
remaining number of years was constant between 1900 and 2015, given that today’s
LE at older ages is longer and is continually increasing. For a more detailed
explanation of prospective age, we refer the reader to the publications of Sanderson
and Scherbov (2008, 2013, 2016), who originated and developed the concept of
Here we give one explanatory example linked to the North Atlantic context. In
1900, a 50-year-old Icelandic woman had 20.4 remaining years to live, while a 50-
year-old Icelandic man had 17.4 remaining years to live (Table 1). We estimated new
prospective ages in 1950, 2000, and 2015 based on these RLEs from 1900. Between
1900 and 2015, the prospective ages of 50-year-old Icelandic men and women rose
by around 15 years. Thus, in Iceland in 2015, 66-year-old women and 67-year-old
men were “the new 50-year-olds” relative to their forebears of 1900. This substantial
72 Population ageing dynamics in the North Atlantic region of the Arctic
The calculated prospective ages in 1950, 2000, and 2015 at which remaining life
expectancy (RLE) is the same as at the age of 50 in 1900, by sex, standard year –
RLE at the age of 50 in 1900 17.4 20.4
1950 60.3 59.8
2000 65.4 64.0
2015 67.4 66.1
Data on life expectancies at diﬀerent ages were obtained from the Human Mortality Database and Statistics
Iceland (for the year 2015). Prospective ages are calculated by the authors.
growth in RLE cannot be ignored when population ageing is measured, as failing to
take prospective age into account may give a false picture of how old the society in
question was in the past, is in the present, and will be in the future.
The prospective indicators in this analysis include the share of people with
an RLE of 15 years or less (Prop RLE 15), the prospective median age (PMA),
the prospective ageing index (PAI), and the prospective old-age dependency ratio
(POADR). The PMA is derived from the life table in which the remaining LE is the
same as the MA in the reference year: in this study, the life table refers to Iceland in
2005. The Prop RLE 15 is calculated as the number of people in age groups in which
the RLE is 15 years or less, divided by the total population. The POADR makes the
share of people older than the RLE of 15 years or less the old-age threshold in the
numerator, which is divided by the number of people between the ages of 15 and
the old-age threshold. The PAI relates the number of people older than RLE 15 or
less to the number of children between the ages of zero and 14. More information
on the computation techniques used for prospective indicators can be found in the
publications of Sanderson and Scherbov from 2008 to 2016 (see references).
Prospective indices consider changes in longevity, and refer to a ﬂoating number
of prospective years to live, as denominated in Table 1, instead of a constant number
of chronological years. For the chronological indices, a 60-year age threshold is used
in the calculations. It may sound artiﬁcial to consider all people older than age 60
as old. This population group is highly heterogeneous and is largely still productive
in labour market terms. Moreover, retirement ages can be well above 60; in Iceland,
for example, the oﬃcial retirement age is 67 (Hansen et al. 2012, 18). However, this
analysis focuses on demographics rather than on labour force analysis, and 60 is
chosen as a reasonable basis for cross-regional comparisons. As the age threshold
of 60 is used in a pan-Arctic study on ageing (data available in Emelyanova ),
relying on this threshold also facilitates numerical comparisons on ageing between
the North Atlantic and the rest of the Arctic region. Regardless of whether we use
Anastasia Emelyanova and Arja Rautio 73
an age threshold of 60 or 65, the trends in ageing dynamics measured will show the
2 Results and discussion
The North Atlantic region has a population that is already out of balance in
many respects (Rauhut et al. 2008; Hansen et al. 2011, 2012a, 2012b; Roto et al.
2014). The complex interrelations of migration, natality, death rates, and family
choices suggest that the size and the characteristics of this population are changing
dramatically, and that policy regulation is needed. Without entering into a broader
discussion of these issues, it is important that we mention several intraregional
demographic features of the area under study: (i) North Atlantic reproduction rates
are comparatively high, but have been declining in recent decades due to changes
in mortality and fertility; and (ii) migration heavily distorts the birth-death balance,
and leads to particularly intense societal ageing trends in rural and remote areas.
These areas are ageing rapidly because large shares of young people (notably young
women) leave to ﬁnd employment, marry, and further their education in the capitals
and larger cities of the North Atlantic, including Nuuk, Reykjavik, Akureyri, and
orshavn. Below, we identify the scope of ageing heterogeneity and cross-territorial
trends for recent decades in the North Atlantic setting.
2.1 Patterns of longevity at a later age
Longevity is a powerful driver of population ageing. Greenland’s average LE was
lower than that of Iceland and the Faroe Islands at the beginning of the 1980s,
which caused a noticeable cross-territorial gap. In 2015, the populations of Iceland
and the Faroe Islands had a life expectancy of 82 years, or eight years more than
the population of Greenland (Statistics Iceland 2016; Statistics Faroe Islands 2016;
Statistics Greenland 2016). In fact, Icelandic men have a LE at birth (81 years)
that is among the highest in the world, and that is two years higher than the LE
of Faroese men. The life expectancies of Icelandic females and males have been
steadily converging due to male “leapfrogging” (Statistics Iceland 2016).
Alternative measures of ageing use the expected time to death derived from life
tables. Lutz and colleagues have argued that the traditional LE at older ages should
be complemented with estimates of how many people with an RLE of 15 years or
less there are in the population, as these people are more likely to be dependent
on public services than the population aged 60 or 65+(Lutz et al. 2008). While
life expectancy has increased across the region, it has grown to varying degrees at
diﬀerent ages (Table 2). Longevity has clearly risen more slowly at later ages than
at birth. For example, an average increase of 5–8 years in LE at birth translates into
an average increase of just 1–4 years at the ages at which RLE is equal to 15 (Age
RLE 15). In recent decades, the growth in the age at which RLE 15 is reached has
74 Population ageing dynamics in the North Atlantic region of the Arctic
been insigniﬁcant among Greenlandic men and women, but has been particularly
strong in the Faroe Islands due to recent increases in longevity, and possibly due to
improvements in the national health care system.
Another indicator that directly reﬂects changes in longevity is the median age
(MA). If we compare the MA standardised for expected RLE (prospective) with
traditional (retrospective) LE, we see that the MA provides more accurate estimates
of changes in the length of a population’s life and health. The MA in the North
Atlantic has been gradually rising for both sexes, and has increased by almost
10 years over the past three decades. In 2013, the MA of the region’s population
reached 36 years, or four years lower than the Danish MA. In 1980, by contrast,
the MA of the North Atlantic population was eight years lower than the Danish
MA. While the cross-territorial diﬀerences have diminished, the gender gap in the
MA has been consistent throughout the period, with women having an MA that was
2–3 years higher than that of men in all territories.
The median age has increased due to two trends that have intensiﬁed since
the end of the 1960s: decreasing birth rates and increasing LE. While the birth
rates in the North Atlantic have declined slightly less than in the Nordic region,
and considerably less than elsewhere in Europe, these decreases have still been
pronounced. Whereas the TFR (total fertility rate) in Iceland rose as high as four
in 1964, it had fallen to 1.8 by 2015 (Statistics Iceland 2016). In Greenland, the
TFR decreased from 2.7 in 1970 to 2.0 in 2015 (Statistics Greenland 2016). In the
Faroe Islands, the TFR declined from 3.4 to 2.4 over the same period (Statistics
Faroe Islands 2016). Deviations from the levels of fertility attained will also have
large eﬀects on population change. According to United Nations forecasts, Iceland
and the Faroe Islands will have TFRs of 1.8 in 2050; while according to the authors’
projections, the TFR in Greenland will grow to 2.2 by 2050.
Future birth rates could also be aﬀected by national policies, and, to some extent,
by abortion rights. In Greenland, women have full access to abortion services; while
in the Faroe Islands and Iceland, access to abortion is restricted to certain cases
of medical or social necessity. The abortion rate in Greenland grew substantially
over several decades prior to the 2000s, whereas in the other two territories, where
abortions are restricted, the rates have been low (Johnston 2016). However, there
have been no drastic changes in abortion rates in the recent past, possibly because
more information is being made available about contraception and family planning.
The question of why these policies have aﬀected fertility rates at the local level needs
to be further investigated. It has been suggested that the North Atlantic is entering
the ﬁnal stage of the Demographic Transition model (from high to low births and
deaths). Additional changes in fertility at the local level could be caused by higher
educational attainment, ﬁnancial uncertainties and crises in local economies, intense
urbanisation, the labour market behaviour of higher educated women, and other
Conventional median ages have been compared to PMAs, which are calculated
on the basis of period life tables and adjusted to the respective changes in life
expectancy. Just as economic analysts compute the outputs of various countries
Anastasia Emelyanova and Arja Rautio 75
Longevity indicators in Denmark and the North Atlantic region by sex (male/female), 1980/2012
Denmark Greenland The Faroe Islands Iceland
1980 2012 1977–1981 2008–2012 1985–1986 2011–2012 1980 2012
Life expectancy at birth 71.4/77.4 78.0/81.9 60.3/67.7 68.7/73.4 72.1/79.4 79.6/84.6 73.5/79.5 80.8/83.9
Life expectancy at age 60 17.1/21.6 21.6/24.2 15.6/17.9 16.4/18.6 18.3/21.9 22.5/24.9 19.3/23.0 23.4/25.4
Age RLE 15 63.1/68.6 67.9/71.1 61.0/64.6 61.9/64.7 64.5/68.7 69.2/72.1 65.8/70.4 69.0/72.4
Median age 33.0/35.3 39.9/41.7 24.7/22.5 35.1/32.0 28.4/28.3 37.6/39.3 26.3/27.3 34.7/35.9
Prospective median age 43.5/36.9 45.1/39.5 37.1/33.5 42.8/40.5 34.5/33.0 36.9/38.7 31.9/30.9 33.9/34.5
“Age RLE 15” is the age at which RLE (remaining life expectancy) equals 15 years. Age RLE 15 data are calculated for a one- to ﬁve-year (Greenland) average. All
median age data are provided for the cut-oﬀs at 1980 and 2012.
Data on life expectancy at birth, at age 60, and at the median age (except for Denmark) are provided by Statistics Denmark, Statistics Greenland, Statistics Faroe
Islands, and Statistics Iceland.
76 Population ageing dynamics in the North Atlantic region of the Arctic
using a standard currency, demographers calculate the PMAs of various territories
using a common life table as a reference (see Sanderson and Scherbov 2008). In
our study, we matched the MA to one standard life table: that is, the life table that
refers to Iceland in 2005. We identiﬁed the ages at which the RLE was the same
as at the MA in the indicated year for the same territory. Appendix 1 contains the
resulting calculations for the PMAs. A diﬀerent picture of ageing emerged when
we applied this technique: while the PMA was growing, it was increasing at half
the pace of the MA: over three decades, the PMA rose 3–4 years (males–females),
while a commonly used MA increased 8–9 years. Across the region, the PMA grew
most rapidly in Greenland, where it increased 6.4 years in the 1980–2010 period.
Over the same period, the PMA grew only 2.5 years in the Faroe Islands and in
2.2 Age structure
In the North Atlantic, cohort eﬀects as well as socioeconomic particularities
continue to have a substantial impact on the development of the population age
structure. One concern is that an ageing population may lead to increased demand
for public services. Small island communities of the North Atlantic face many
challenges related to ageing. For instance, because they have diﬃculties recruiting
qualiﬁed specialists in gerontology, geriatrics, and other medical specialties,
patients are sometimes sent for treatment to Denmark or Iceland. In addition to
being expensive, this system can impose hardships on older people. Thus, further
developments in local infrastructure and regional cooperation in the provision of
medical care for the elderly are needed.
According to Gløersen et al. (2005: 64), demographic changes are “the most
comprehensive synthetic indicator of economic and social dynamism”, and have
a big impact on age structure. The North Atlantic experienced strong population
growth in the last century. At the beginning of the 20th century, approximately
100,000 people lived in the Faroe Islands, Greenland, and Iceland. Within one
hundred years, the population of Greenland had increased more than ﬁvefold, the
population of Iceland had increased fourfold, and the population of the Faroe Islands
had more than tripled. These growth rates are tremendous when compared with the
general population declines that were occurring in Europe and in other regions of
the Arctic. However, from the 1990s onwards, this growth trend reversed, and turned
into a trend towards “thinning out societies” (Aasbrenn 1989).
Currently, the populations of the North Atlantic region have the lowest propor-
tions of elderly people and the highest shares of children and young people in the
Nordic country group. While Greenland has the youngest population by any mea-
sure (Foss and Juvkam 2005), its birth rates have been declining in the period stud-
ied. In Iceland, the share of young people in the population has also increased. Thus,
compared with the rest of the Europe, the region has a relatively young labour force.
Anastasia Emelyanova and Arja Rautio 77
Danish and North Atlantic proportions of the population with a remaining life
expectancy of 15 years or less (Prop RLE15-), sexes combined, (%) of total
Dk: Denmark, Gl: Greenland, Fa: the Faroe Islands, Ic: Iceland. The data are the authors’ calculations based
on population counts and mortality numbers provided by Statistics Denmark, Statistics Greenland, Statistics Faroe
Islands, and Statistics Iceland.
When the changes in broader population age groups (youth, adults, and the
elderly) are tracked, clear shifts can be seen over the last three decades (Figure 2).
The shrinking number of children and teenagers is a direct result of the tendency
among young people in the North Atlantic to have fewer children. Urbanisation
plays a large role in this shift. As Hansen et al. (2012) observed, the more rural
areas of the Faroe Islands are experiencing a greater reduction in the proportion of
the young people in the population than the rural areas of Greenland. Greenland
has experienced a similar decline in its capital region, whereas in Iceland, there has
been a decrease in all parts of the country.
Because of the relative lack of educational and employment opportunities in the
North Atlantic, migration is a signiﬁcant accelerator of population ageing in the
region, causing the shares of young and working-aged people in the population to
shrink. Although the North Atlantic is remote and travel is expensive, the people
are mobile (Hansen et al. 2011), in part because Nordic citizens are permitted to
move freely within the Nordic region under regional labour and educational policies
(Dustmann and Albrecht 2011). There is, however, a noticeable gender discrepancy
in the numbers of young people emigrating from small settlements, as women are
more likely than men to permanently leave their home island communities and the
region itself (Rasmussen 2005), taking with them their labour skills and potential
oﬀspring. More women than men from Greenland, and to a lesser degree from
the Faroe Islands, have moved to Denmark to enter into interethnic marriages with
78 Population ageing dynamics in the North Atlantic region of the Arctic
Danes. Some older people from Greenland and the Faroe Islands move to Denmark
after retirement, which tends to increase the share of people aged 60+in the Danish
population. Figure 2 shows that Denmark had the largest share of people aged 60+
in the population in 1980 (19%) and in 2010 (23%).
Denmark also has the smallest share of children in the population. The North
Atlantic regions used to perform better in this regard, and more data can be seen in
Appendix 1. At the same time, an intense wave of out-migration of young people
and children in recent years has caused the islands’ populations to age even further.
If we compare 1980 to 2010, we see that the proportion of working-aged (15–59)
and older (60+) people in the population increased by 0.9% and 5.5%, respectively,
in the Faroe Islands, by 2.1% and 5.3% in Greenland, and by 3.4% and 3.3% in
2.3 Proportion of older people
Focusing speciﬁcally on the proportion of older people in the age structure, we see
that the chronological age-based methods indicate that the sex balance has changed
in recent years. From 1980 to 2013, the share of men aged 60+in the population
grew from 4.5% to 12% in Greenland. In Iceland, this share had reached 12% by the
1980s. The number of Icelandic men aged 60+increased by almost one-third over
this period. In 2013, the Faroe Islands had the highest share in the region of men
aged 60+in the population (20.6%). Among women aged 60+, the fastest changes
occurred in the Faroe Islands, where the share of older women in the population
reached 23.2% in 2013. The share of women aged 60+in the population reached
somewhat lower values in Iceland (19%) and in Greenland (11.3%) in 2013. In
Greenland, the shares of men and women aged 60+were roughly the same in 1980,
but by 2013, older men outnumbered older women. This is an unusual situation
given that older women outnumber older men in most parts of the world. The
explanations for this gender imbalance include the high rates of out-migration of
women from the island since the beginning of the 1990s and the predominance of
male-oriented employment sectors. Rauhut et al. (2008) also observed that the share
of people aged 60+in the population is relatively low in Greenland due to the shorter
LE and the fact that many pensioners choose to settle in Denmark.
The related Prop 60+measure refers to the proportion of people considered old
when the average remaining lifespan in their age group is less than 15 years. Figure 3
illustrates the evolution of Prop RLE 15 proportions in the 1980–2010 period. When
we weight the RLE of all age groups with the proportions of people belonging to
those age groups in the population, the dynamics of population ageing are altered.
We see changes for both sexes: Prop RLE 15 declines slightly in every territory,
except in Greenland due to its lower LE. At the end of the period, the diﬀerences
between the North Atlantic territories diminish noticeably, and the ratios, including
the OADRs, start to converge. Hence, the historical ageing trends look diﬀerent than
they did when we were using the conventional deﬁnition of elderly people.
Anastasia Emelyanova and Arja Rautio 79
Danish and North Atlantic proportions of the population with a remaining life
expectancy of 15 years or less (Prop RLE 15) by sex, (%) of total population
The data are the authors’ calculations based on population count and mortality numbers provided by Statistics
Denmark, Statistics Greenland, Statistics Faroe Islands, and Statistics Iceland.
2.4 Dependency ratios
The old-age dependency ratio (OADR), a common measure of ageing, corresponds
to the ratio between the number of older people and the total working-aged
population (aged 15–59). The current OADRs were calculated based on the
knowledge that the complementary measure, the young age dependency ratio, is
higher in Iceland, Greenland, and the Faroe Islands than in other Nordic territories.
For the middle-share ratios, there is, as mentioned above, an upwards trend in the
migration of young, educated women to other economically prosperous areas and
countries (Rasmussen 2011).
Over the study period, the highest OADR load for both sexes was in the Faroe
Islands, followed by in Iceland. Greenland had the lowest OADR due to the
relatively large share of young people in the age structure of its population. The
speed of change of the ratio diﬀered over the period, with the fastest growth
occurring in the Faroe Islands, particularly among women. The OADR grew half
80 Population ageing dynamics in the North Atlantic region of the Arctic
as quickly in Greenland and in Iceland as in the Faroe Islands among women.
Greenland’s male population experienced relatively fast OADR growth. Topping
the whole North Atlantic region was Denmark, where the indicator was two to three
times higher than in Greenland.
Given this pronounced growth in old-age dependency, as measured convention-
ally, over just over three decades, local governments and stakeholders may raise
concerns about the sustainability of current welfare systems. However, it must be
taken into account that, as people age, they tend to gain healthy years, and thus do
not automatically become a burden on the health care and social welfare systems
after the age of 60. These dependency ratios change when the prospective OADR
is calculated, as both the numerator and the denominator are aﬀected by the old-
age threshold. In the prospective OADR, the old-age threshold is connected to the
ﬂoating LE, rather than being set at age 60. Increases in the threshold age (RLE 15)
reduce the number of people considered old, and hence increase the share of people
of working age. Figure 4 illustrates that the dependency ratios are much lower using
this approach than using the chronological approach. The POADRs for both sexes
and in all areas increase less rapidly than when the traditional OADR is applied.
Moreover, the ﬁgure shows that the POADR has been slowly decreasing in most of
the North Atlantic since the 1990s, whereas the OADR has been growing.
By replacing the number of people aged 60 years or older in the calculations with
the number of people older than the speciﬁc old-age threshold, we provide a new
type of evidence regarding societal dependency in old age, adjusted to longevity
and health transformations. This exercise showed a decrease in dependency in
Denmark and the Faroe Islands for both sexes, and for males in Iceland. One
exception is Greenland, where the POADR has been growing. This trend is
related to the accelerating ageing and the relatively early stage of the health and
demographic transition of the Inuit population (“natives born in Greenland” in
Statistics Greenland) who made up 89% of the total population in Greenland in
2015 (Statistics Greenland 2016).
2.5 Ageing indices
It is important to address the ageing index (AI) when analysing changes in age
structure, as it is greatly inﬂuenced by survival and fertility rates. The composite AI
shows the interrelations between the old and the young age groups, and estimates
how fast a population is ageing by calculating the number of people aged 60+for
every 100 young people under age 15 (the elder-child ratio). Figure 5 shows that
the AIs generally increased among the populations in the North Atlantic between
1980 and 2010. The Faroe Islands had the highest AI for both sexes, followed by
Iceland, with Greenland lagging considerably behind. The AIs of the Faroe Islands
and Iceland grew rapidly in recent years due to the sharp rise in the proportion of
the population with higher education, which is an important factor in falling birth
rates, among other things. When comparing chronological and prospective AIs, it is
Anastasia Emelyanova and Arja Rautio 81
Chronological and prospective old-age-based dependency ratios in Denmark and the
North Atlantic region, sexes combined, 1980–2010
Denmark Greenland Faroe
OADR, both sexes
Denmark Greenland Faroe
POADR, both sexes
OADR: Old-age dependence ratio, POADR: Prospective old-age dependency ratio. The data are the authors’
calculations based on mortality numbers provided by Statistics Denmark, Statistics Greenland, Statistics Faroe
Islands, and Statistics Iceland.
interesting to observe that the PAI values rose more slowly than the AI values, and
that there were also some downwards trends, such as in the Faroese population after
1995. These developments led to a convergence (particularly among males) in the
PAI rates by 2010 across the three North Atlantic territories (Figure 5).
The ageing variations across the North Atlantic considered in this study are
determined by broad demographic drivers:
82 Population ageing dynamics in the North Atlantic region of the Arctic
Chronological and prospective old-age-based ageing indices in Denmark and the
North Atlantic region, sexes combined, 1980–2010
1980 1990 2000 2010
AI, both sexes
1980 1990 2000 2010
PAI, both sexes
Faroe Islands Iceland
Faroe Islands Iceland
AI: Ageing index, PAI: Prospective ageing index. The data are the authors’ calculations based on population
counts and mortality numbers provided by Statistics Denmark, Statistics Greenland, Statistics Faroe Islands, and
•Iceland and the Faroe Islands have one of the highest life expectancies in the
world (82 years in 2015) and the lowest mortality levels in the world, whereas
life expectancy in Greenland lags eight years behind (74 years in 2015).
Anastasia Emelyanova and Arja Rautio 83
•The TFR has historically been higher in the North Atlantic than in nearly
every other country in Europe, but as was noted in the results and discussion
sections, it has been falling since the end of the 1960s: i.e., the average TFR
across the three territories declined from three in 1970 to two in 2015. This
value is around the replacement level and above the Danish national average.
•A powerful ageing driver within this geography is the meaningful out-ﬂow
of college-bound youth (representing 15% of youth emigration in the age
group 20–24 in the Faroe Islands in 2015), and of educated people in older
age groups, especially from rural areas (Statistics Faroe Islands 2016). This
trend started to accelerate in the 2000s, and was accompanied by large waves
of out-migration after the ﬁshery/banking crises. In such conditions, ageing
can exacerbate the challenges faced by rural and ﬁshery communities, acting
in concert with factors such as limited access to public services and transport,
and decreasing income levels.
Following the study aims outlined in the introductory part of the paper, and
based on the data summarised in Appendix 1, the results shed light on the oldest
and the youngest, and the fastest and the slowest areas of ageing. The traditional
measurements indicate that there has been an explicit upwards trend in ageing in all
of the North Atlantic territories except in Denmark, where the MA and the OADR
decreased in 1990–2000 and the AI decreased until 2010. In 2010 the most advanced
ageing was found for the Faroese population, with Greenland lagging far behind.
However, from 1980 to the present, Greenland (particularly the male population) has
been experiencing the fastest ageing according to chronological measures, with the
indices’ values increasing two- to threefold. The trends in ageing in Greenland are
also distinct in terms of sex diﬀerences, with older men outnumbering older women,
which is rare worldwide. Meanwhile, the traditional indicators showed that Iceland
has been ageing less rapidly, with its rates occupying a middle position between
those of its two neighbours. Moreover, according to the chronological indices, the
ageing rates in the North Atlantic territories have not been converging, but have
instead been developing in parallel, while keeping the same ranking throughout the
The results of the prospective ageing indicators diﬀered markedly from those
of the chronological indicators. The prospective indicators showed that the ageing
rates of the North Atlantic territories have been converging, and had reached
almost the same degree of ageing by the 2010 observation, particularly among men.
These ﬁndings diverge not just from those of traditional measurements of ageing
dynamics, but from ﬁndings for some other regions of the Arctic, such as for the
North American Arctic (Alaska and the Canadian North), where a high degree
of cross-territorial divergence has been found (Emelyanova 2015). The ranking of
the territories also changed. While the Faroe Islands had the highest chronological
ageing rates, Greenland had the highest prospective estimates in 1980 and 2010 (e.g.,
PMAs). The prospective approach also showed a slower pace of ageing at several
points in time. For Iceland and the Faroe Islands, this process even appeared to be
84 Population ageing dynamics in the North Atlantic region of the Arctic
reversed in the most recent decade, signifying a so-called population rejuvenation.
A rejuvenation trend was found when the POADR rates (2000s) were applied to
these two territories, and the Prop RLE 15 was shown to have declined for males in
Iceland and in the Faroe Islands over the 1980–1990 period. Similar trends in the
PAI, the PMA, and the Prop RLE 15 after the 2000s were found among Faroese
women. Almost all of the indices of prospective ageing for Denmark were shown to
The future path of ageing in the North Atlantic communities needs to be
continually monitored and forecasted, as smaller populations can experience greater
ﬂuctuations and more abrupt changes in terms of age structure and population size
than larger populations. The way ageing is measured also needs to be taken into
account, as we have shown for the case of the North Atlantic in this study. The
prospective measurement of ageing indicates that societal ageing is not developing
as quickly or as linearly as the chronological approach has shown; thus, our
understanding of the consequences of ageing and our responses to this phenomenon
may need to be reconsidered.
Aasbrenn, K 1989. Uttynningssamfunnet. Det demograﬁsk uttynnede - men ikke avfolkede -
utkantsamfunn in. Tidskrift for samfunnsforskning 30(5/6): 509–519.
Danilova, R., E. Golubeva and A. Emelyanova 2011. Demographic aspects of the social safety
of the elderly population in the Barents Euro-Arctic region. Advances in Gerontology
Denton, F. and B. Spencer 2000. Some demographic consequences of revising the deﬁnition
of ‘Old’ to reﬂect future changes in life table probabilities. SEDAP Research Paper
2000–22. Hamilton, Canada: McMaster University.
Dustmann, C. and G. Albrecht 2011. Migration and education. Discussion Paper 2011–
11. Norface Research Programme on Migration, Department of Economics. London:
University College London.
Emelyanova, A. and A. Rautio 2012. Ageing population in the Barents Euro-Arctic region.
Journal of European Geriatric Medicine 3(3): 167–173.
Emelyanova, A. and A. Rautio 2013. Perspectives for population ageing in the Russian North.
Journal of Population Ageing 6(3): 161–187.
Emelyanova, A. 2015. Cross-regional analysis of population aging in the Arctic. Academic
doctoral dissertation. Acta Universitatis Ouluensis D Medica 1326, Juvenes Print
Tampere, 2015, pp. 1–94. http://herkules.oulu.ﬁ/isbn9789526210049/isbn9789526210049.
pdf, accessed on 23 August 2016.
Emelyanova, A. and A. Rautio 2016. Population diversiﬁcation in demographics, health, and
living environments: the Barents Region in review. Nordia Geographical Publications
Yearbook 2016. Forthcoming.
Foss, O. and D. Juvkam 2005. Patterns of demographic ageing and related aspects in the
northern peripheries. Nordregio Report 2005–2. Stockholm: Nordregio.
Anastasia Emelyanova and Arja Rautio 85
Gløersen, E., A. Dubois, A. Copus and C. Sch
urmann 2005. Northern peripheral, sparsely
populated regions in the European Union. Nordregio Report 2005–04. Stockholm:
Golubeva, E. and R. Danilova 2011. Improving the social safety of elderly population
in the system of social work on sparsely populated areas of European North of Russia.
Medico-Biological and Socio-Psychological Problems of Safety in Emergency Situations 4:
Hamilton, L. and A. Mitiguy 2009. Visualizing Population Dynamics of Alaska’s Arctic
Communities. Arctic 64(4): 393–398.
Hansen, K., R. Rasmussen and J. Roto 2011. Demography in the Nordic countries – a
synthesis report. Nordregio Working Paper 2011–09. Stockholm: Nordregio.
Hansen, K., R. Rasmussen, J. Roto and C. Fredricsson 2012a. Megatrends in the Arctic – new
inspiration into current policy strategies. Nordregio Working Paper 2012–08. Stockholm:
Hansen, K., R. Rasmussen and J. Roto 2012b. Nordic perspectives on demography – a
background report for the project on coastal societies and demography. Nordregio Working
Paper 2012-12. Stockholm: Nordregio.
Johnston, R. 2016. Historical abortion statistics. http://www.johnstonsarchive.net/policy/
abortion/index.html, accessed on 27 September 2016.
Larsen, J. N. and G. Fondahl, eds 2014. Arctic human development report. Regional processes
and global linkages. Copenhagen: Nordic Council of Ministers.
Leeson, G. 2002. The changing face of the population in Europe. Nordregio Working Paper
2002–02. Stockholm: Nordregio.
Lewis, J. 2013. The future of successful ageing in Alaska. International Journal of
Circumpolar Health. doi.org/10.3402/ijch.v72i0.21186.
Lutz, W., W. Sanderson and S. Scherbov 2008. The coming acceleration of global population
ageing. Nature 451: 716–719.
Moore, E. and M. Pacey 2004. Geographic Dimensions of Ageing in Canada, 1991–2001.
Canadian Journal on Ageing/La Revue Canadienne du Vieillissement 23(suppl.): 5–21.
om, L. and J. Roto 2013. A Nordic ‘Agequake’? Population Ageing in
Nordic Cities and Regions. Nordregio News Issue 2013–03, Stockholm: Nordre-
Ageing–Challenge-and-Opportunity/Context/, accessed on 24 August 2016.
Nummela, O., M. Sepp
anen and A. Uutela 2011. The eﬀect of loneliness and change in
loneliness on self-rated health (SRH): a longitudinal study among ageing people. Archives
of Gerontology and Geriatrics 53(2): 163–167.
Rasmussen, R. 2005. Analysis of hunters and ﬁshermen in Greenland. Grønlands
Hjemmestyre - Direktoratet for Fangst og Fiskeri og NORS – Nordatlantiske
Regionalstudier. Roskilde: Roskilde Universitets Center.
Rasmussen, R., ed. 2011. Megatrends. Copenhagen: TemaNord, Nordic Council of Ministers.
Rauhut, D., R. Rasmussen, J. Roto, P. Francke and S.
Ostberg 2008. The demographic
challenge to the Nordic countries. Nordregio Working Paper 2008–01. Stockholm:
Roto, J., J. Grunfelder and L. Rispling, eds. 2014. State of the Nordic region 2013. Stockholm:
86 Population ageing dynamics in the North Atlantic region of the Arctic
Roto, J. and R. Rasmussen 40 years of population change in West Norden. http://www.
accessed on 24 August 2016.
Sanderson, W. and S. Scherbov 2008. Rethinking age and ageing. Population Bulletin 63(4):
Sanderson, W. and S. Scherbov 2010. Remeasuring Ageing. Science 329(5997):
Sanderson, W. and S. Scherbov 2013. The characteristics approach to the measurement of
population ageing. Population and Development Review 39(4): 673–685.
Scherbov, S. and W. Sanderson 2016. New approaches to the conceptualization and
measurement of age and aging. Journal of Aging and Health 28(7): 1159–1177, accessed
on 16 December 2016.
Statistics Faroe Islands 2016. http://www.hagstova.fo/en, accessed on 23 August 2016.
Statistics Greenland 2016. http://www.stat.gl/default.asp?lang=en, accessed on 23 August
Statistics Iceland 2016. http://www.statice.is/, accessed on 23 August 2016.
United Nations. 2015. World Population Ageing 1950–2050. New York: Population Divi-
sion DESA. (ST/ESA/SER.A/390). http://www.un.org/en/development/desa/population/
publications/pdf/ageing/WPA2015 Report.pdf, accessed on 16 December 2016.
Wilson, D., J. Osei-Waree, J. Hewitt and A. Broad 2012. Canadian provincial, territorial,
and federal government aging policies: A systematic review. Advances in Aging Research
World Health Organization. 2011. Life expectancy ranking. http://www.who.int/gho/
mortality burden disease/life tables/situation trends text/en/, accessed on 21 December
Population ageing indicators for the North Atlantic region by sex, 1980–2010
Year Sex MA PMA Prop Prop OADR POADR AI PAI Age 15
60+RLE 15 ×100 ×100 ×100 ×100
1980 Males 24.72 37.13 4.51 4.17 6.76 6.14 15.67 13.32 61.02
1990 28.53 40.49 5.45 5.81 7.74 8.26 22.60 24.42 59.23
2000 32.56 42.53 7.34 7.40 10.96 11.09 28.52 28.54 59.74
2010 35.08 42.84 10.85 8.80 16.12 12.78 49.71 39.45 61.90
1980 Females 22.52 33.51 6.54 4.50 10.55 7.22 20.74 13.55 64.63
1990 26.81 36.97 7.54 4.78 11.59 7.00 27.59 17.83 64.73
2000 29.59 41.69 8.66 7.10 13.84 11.09 30.13 24.64 62.54
2010 32.01 40.52 10.59 7.23 16.12 10.57 44.61 29.73 64.68
Anastasia Emelyanova and Arja Rautio 87
Year Sex MA PMA Prop Prop OADR POADR AI PAI Age 15
60+RLE 15 ×100 ×100 ×100 ×100
1980 Males 26.30 31.95 12.24 8.25 20.53 12.98 43.51 29.34 65.80
1990 29.38 32.52 13.26 8.08 21.63 12.16 52.11 31.77 66.89
2000 32.13 33.12 13.83 7.83 22.16 11.45 58.12 32.90 68.84
2010 34.27 33.92 15.73 8.07 24.94 11.41 74.12 38.03 68.99
1980 Females 27.34 30.94 14.63 7.36 25.17 11.25 53.74 27.03 70.41
1990 30.25 32.69 15.74 7.70 26.36 11.37 64.09 31.37 70.85
2000 32.48 34.97 16.27 8.41 26.73 12.23 71.20 36.78 71.06
2010 35.31 34.54 17.78 8.15 28.87 11.45 86.12 39.50 72.39
1985 Males 28.39 34.54 13.70 10.50 23.48 16.40 49.05 41.32 64.51
1990 30.39 35.46 14.04 10.78 22.76 16.60 57.75 44.34 64.24
2000 33.96 35.34 15.73 9.18 25.88 13.63 67.05 39.13 68.17
2010 36.71 36.63 18.85 8.50 31.70 12.18 87.02 39.25 70.00
1985 Females 28.30 33.02 15.74 9.23 28.28 14.17 54.97 35.94 68.74
1990 31.06 33.74 17.53 9.71 30.21 14.75 71.76 39.75 69.22
2000 35.15 35.41 19.58 10.09 34.84 15.37 80.85 41.68 72.14
2010 38.18 35.34 21.58 8.68 38.32 12.55 97.58 39.26 74.46
1980 Males 33.00 43.51 17.16 14.12 28.14 21.89 78.40 66.17 63.12
1990 35.71 45.64 17.74 14.28 27.51 21.02 99.65 80.21 63.69
2000 37.23 45.23 17.34 12.29 27.29 17.91 90.76 64.29 65.25
2010 39.60 45.09 21.33 11.19 35.57 15.96 114.07 59.84 67.89
1980 Females 35.26 36.89 21.48 11.75 36.94 19.05 105.50 59.18 68.65
1990 38.42 39.86 22.97 14.04 37.97 20.21 139.12 85.01 68.97
2000 39.39 39.78 22.03 13.25 36.56 19.19 124.28 74.75 69.50
2010 41.28 39.47 24.73 11.54 42.80 16.27 141.38 66.00 71.14
(prospective) median age, PMA, refers to Iceland in 2005 as a reference year; Prop 60+is the
share of people aged 60 and older; Prop RLE 15 is the share of people with a remaining life expectancy of 15
years or less; OADR/POADR – (prospective) old-age dependency ratio; AI/PAI – (prospective) ageing index; Age
15 is the age at which the remaining life expectancy is equal to 15. Greenland’s data are provided for a ﬁve-year
average only, as our calculations are based on mortality tables downloaded from Statbank of Statistics Greenland,
provided for a ﬁve-year interval only. The table indicators refer to the periods of 1977–81, 1987–91, 1997–2001,