Chuluun, T., Graham, C., & Myanganbuu, S. (2016). Who is happy in the land of eternal blue sky? Some

insights from a first study of wellbeing in Mongolia. International Journal of Wellbeing, 6(3), 49-70.

doi:10.5502/ijw.v6i3.506

Carol Graham

The Brookings Institution

cgraham@brookings.edu

www.internationaljournalofwellbeing.org

ARTICLE

Who is happy in the land of eternal blue sky? Some

insights from a first study of wellbeing in Mongolia

Tuugi Chuluun · Carol Graham · Sarandavaa Myanganbuu

Abstract: We conducted the first extensive study of wellbeing in Mongolia, a country that has

experienced a dramatic transition in both its economy and polity in recent decades. We found that

most of the standard determinants of wellbeing were no different in Mongolia than they are for

most countries in the world, despite the unique context and the extreme changes there, with

individual income, health, marital status, and exercise all positively associated with life

satisfaction. The same variables had positive but weaker correlations with our measure of hedonic

wellbeing. When we split our sample into those above and below median income, however, there

were some important differences in the findings. As in many other contexts, stress is negatively

correlated with wellbeing in Mongolia, and the primary triggers were concerns about income,

family, and the living environment. Finally, we found that freedom of expression has an important

positive association with wellbeing in Mongolia, perhaps reflecting the extent to which it is an

island of political freedom compared to its neighbors.

Keywords: subjective wellbeing, Mongolia, community income, stress, freedom of expression

1. Introduction

There is a burgeoning literature on wellbeing around the world, much of which finds consistent

patterns in its determinants in countries and cultures around the world. Many of these patterns

are predictable: income matters to individual wellbeing, but after a certain point other things

such as the incomes of others also start to matter. Health is essential to wellbeing, and stable

partnerships, stable marriages, and social relationships also play a role. Women are typically

happier than men, except in contexts where their rights are severely compromised (Graham &

Chattopadhyay, 2013). And because these patterns are so consistent across diverse countries and

cultures, scholars in the field can control for these factors and explore the wellbeing effects of

phenomena that vary more, such as inflation and unemployment rates, crime and corruption,

smoking, drinking, and exercising, and the nature of public goods, among others (Frey & Stutzer,

2002; Blanchflower & Oswald, 2004; Graham, 2009; Graham, 2008; Helliwell, Layard, & Sachs,

2013). There is also a nascent literature on the causal properties of wellbeing, which finds that

happier people are, for the most part, healthier and more productive (Graham, Eggers, &

Sukhtankar, 2004; DeNeve & Oswald, 2012).

Within this broader framework, we undertook the first extensive survey of wellbeing in

Mongolia, a unique context for a number of reasons. Mongolia, landlocked between China and

Russia, is home to one of the world’s last surviving nomadic cultures, and the country is not only

very remote, but also the least densely populated one in the world after Greenland. Citizens in

Mongolia recently experienced an unusually dramatic transition in the nature of their economy

and political system. After 70 years under a socialist rule, the country made a transition to a

mailto:cgraham@brookings.edu
http://creativecommons.org/licenses/by-nc-nd/3.0/

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 50

market economy in the early 1990s, and recently experienced economic growth rates which were

among the highest in the world (17.2% in 2011), fueled by a mining boom. Mongolia also ranks

number one in the world in terms of having the smallest gender gaps in economic participation

and opportunity, and in health and survival.1

For all of these reasons, one could expect that wellbeing trends in Mongolia might diverge

from the usual patterns that we find elsewhere, and a primary question for us was whether the

basic patterns in the determinants of wellbeing trends would hold in Mongolia. On the one hand,

the basic determinants of wellbeing are fairly consistent in countries around the world (Graham,

2009). On the other, there are some departures for the transition economies in general compared

to the EU and to Latin America, such as a higher correlation between income and wellbeing but

a smaller one for unemployment (this may be due to generous unemployment benefits in some

transitions economies and/or to the increasing importance of the informal economy in others).

The transition economies also display a stronger correlation between freedom to choose and

wellbeing than do those in Latin America (but not compared to Europe) (Graham & Nikolova,

2015).

Because of the detailed and disaggregated nature of the data that we collected through our

survey, we were able to explore additional questions that larger scale, less fine-grained data sets

do not allow for.2 In particular, we focused on the standard determinants of wellbeing and how

they varied, depending on where in the income distribution respondents were, as well as some

additional determinants that are unique to the Mongolian context and available through the

survey. As is increasingly common in the literature, we analyzed two distinct dimensions of

wellbeing – hedonic and evaluative – separately, comparing our findings across these

dimensions in Mongolia to those that we have based on worldwide data (Graham & Nikolova,

2015). These two distinct and measurable dimensions of wellbeing capture different aspects of

human lives.3 The first is hedonic wellbeing, which captures the manner in which individuals

experience their daily lives, the quality of those lives, and their moods (both positive and

negative) during those experiences. The second is evaluative wellbeing, which captures how

people think about and assess their lives as a whole. The latter dimension implicitly includes

eudaimonic wellbeing – how much purpose or meaning people have in their lives – although

there are also aspects of daily experiences which can be purposeful but not pleasurable (such as

reading the same story over and over again to a child) and others which are pleasurable but not

purposeful (such as watching television).

Hedonic wellbeing is typically measured with questions that gauge positive affect on the one

hand (smiling yesterday or happy yesterday, for example) and negative affect (anger or stress

yesterday) on the other. Psychologists emphasize that there is not a simple continuum running

from positive to negative dimensions, as people can experience both at the same time, such as

happiness and stress (e.g., Stone & Mackie, 2013). Evaluative wellbeing, meanwhile, is typically

measured with questions which ask respondents about their satisfaction with their lives as a

whole, or to compare their lives to the best possible life they can imagine.

1 This is according to the Global Gender Gap Index 2012 from the World Economic Forum, which measures gender-

based gaps in access to resources and opportunities around the world rather than the actual levels of the available

resources and opportunities.
2 While the Gallup World Poll has included Mongolia for several years, and the basic questions in our survey are

similar, our survey has, in addition, a number of questions which are tailored to the unique Mongolian context.
3 For a detailed discussion among many scholars in the field and their report for the National Academy of Sciences on

wellbeing metrics and their distinct dimensions, see Stone and Mackie (2013).

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 51

Evaluative wellbeing typically correlates more closely with individual income than hedonic

wellbeing, not least as life course evaluations extend well beyond momentary experiences and

encompass the opportunities and choices that people have in their lives. A nascent body of

research suggests that the dimension of wellbeing individuals value most may be mediated by

their agency and capacity to control their lives (Kahneman & Deaton, 2010; Graham & Nikolova,

2015). Kahneman and Deaton (2010) find that income correlates much more closely with

evaluative than hedonic wellbeing in the United States. The positive correlation between hedonic

wellbeing and income tapers off at roughly USD75k, or, median income, but the association

between income and evaluative wellbeing continues in a log-linear fashion. This suggests that

beyond a certain point, additional income cannot make people enjoy their daily lives more

(although insufficient income is clearly linked to suffering and negative moods), but higher levels

of income offer people many more choices about how to live and what to do with their lives.

Graham and Nikolova (2015) find that individuals emphasize one wellbeing dimension over

the other, depending on their agency and capabilities. Respondents with more means and agency

(e.g., the capacity to make choices over the course that their life takes) tend to emphasize

evaluative wellbeing more, while those with limited means and opportunities tend to emphasize

daily experience more. They also find that income and agency are less important to the wellbeing

of respondents who are at the highest levels of the wellbeing distribution. As noted above,

income and freedom to choose have a relatively high importance for the wellbeing of transition

economy respondents compared to other regions, while unemployment status seems to matter

less. This may be because both differential monetary rewards for effort and skills, and freedom

to choose were sorely lacking in many countries under central planning, and are thus more

appreciated in the post-transition context.

There is a wide literature and extensive debate on the relationship between relative income

and wellbeing. This is, in part, because the effects of inequality on individual welfare – which

seem to partially hinge on comparisons with peers, neighbors, or other relevant cohorts – are

rarely captured by large-scale aggregate measures. In part, it is because inequality signals

different things to different people, depending on the context. Some studies in the transition

economies find that inequality has positive signaling effects rather than negative comparison

effects (Senik, 2004; Cojocaru, 2012); we explore this in the Mongolian context below.

2. Mongolia in transition: The context

Mongolia is one of the most sparsely populated countries in the world, with a population of

around 2.8 million in 2012 and land area of 1.56 million square kilometers (see Table 1 below). It

is also home to one of the world’s last surviving nomadic cultures, with approximately 40% of

the country’s workforce still maintaining a nomadic lifestyle and herding livestock. With a GDP

per capita of around $3,600 and a GDP of $10.32 billion in 2012, Mongolia falls into the lower

middle income category, according to the World Bank classifications.

After 70 years under socialist rule, the country experienced a relatively peaceful transition

from a centrally planned socialist economy to a market economy, following the fall of the Soviet

Union, and democratic government quickly emerged. Currently, the country is governed by a

mixed presidential-parliamentary system, and despite political crises from time to time,

Mongolia is considered free and relatively stable with little violence. Many, in fact, highlight how

on the Freedom House global map, Mongolia appears as an island with its “free” status

surrounded by other nations rated as “not free.”

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 52

Table 1. Key statistics for Mongolia, 2012

Average

Mongolia

Lower

Middle

Income

Global

GDP (current $ billion) 0010.32

Population (million) 0002.79

Human Development Index rank 103/187

Poverty at national poverty line (% of population) 0027.40

Gini coefficient, 2003-2012 0036.52

Population density (People per sq.km of land area) 0001.80 0121.40 000054.29

GDP per capita (current $) 3,691.05 1,998.64 10,437.76

GDP growth (%) 0012.40 0004.55 000002.26

Life expectancy at birth (years) 0067.34 0066.21 000070.78

Unemployment (% of total labor force) 0005.20 0005.29 000005.95

Adult literacy rate (% of people ages 15+), 2010 0098.26 0070.58 000084.29

Female labor force participation rate (% of female

population ages 15+)

0056.10 0036.28 000050.20

Ratio of female to male labor force participation rate

(%)

0081.54 0048.11 000068.25

Source: World Development Indicators, UNDP

The transition period, however, also brought deep recession, hyperinflation, and food shortages

– common in many other post-Soviet countries. Overall, the transition economies experienced

substantial drops in both income and wellbeing during the change from centrally planned to

market economies, with wellbeing demonstrating a U-shaped curve over time: falling

dramatically in the initial transition years and then recovering as economies stabilized and grew.

The extent to which wellbeing recovered to its pre-transition levels, though, depends on

particular countries and the state of their economies. When split into specific domains,

meanwhile, wellbeing recovered more in pecuniary areas – such as financial satisfaction – than

it did in others, such as health satisfaction and satisfaction with family life. Given the dramatic

changes that occurred in most post-Soviet countries’ social welfare systems, this is not a surprise

(e.g., Easterlin, 2009).

In recent years, however, the Mongolian economy has been growing rapidly, fueled by a

mining boom. The economy grew by 17.2% in 2011 and 12.4% in 2012, among the highest in the

world and expected to grow at a double-digit rate over the period from 2013 to 2017, according

to the World Bank. Although the rapid economic growth raised the GDP per capita from $514 in

2005 to $3,691 in 2012, Mongolia ranked 103rd out of 186 countries in 2012 in terms of human

development, according to the United Nations Human Development Index. In addition, about

one-third of the population lives in poverty, according to Mongolian national statistics. The Gini

coefficient for the 2003-2012 period was a relatively low 36.52, as shown in Table 1 above. Hence,

it faces many challenges that are common to transition economies, as well as many that are

unique to the country and its people.

Some of the notable features of the country and its people include a high adult literacy rate

of 98.26% and high female labor force participation. As noted briefly above, gender gaps are

relatively small in Mongolia, and labor force participation as a percentage of the female

population and the ratio of female to male labor force participation are higher than the global

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 53

and lower middle income country averages (see Table 1 above), and Mongolia ranks number one

in the world in terms of small gender gaps in the two categories of economic participation and

opportunity, and health and survival, according to the Global Gender Gap Index 2012. Moreover,

approximately half of the population follows Tibetan Buddhism, and about 40% of the people do

not practice any religion, according to various national statistics, in part due to the ban that was

placed on religious practice under the communist government. Shamanism/Tengrism, which

was the dominant religion historically, is also experiencing a modern revival, and Mongolians

commonly refer to their homeland as “the Land of Eternal Blue Sky”.

As for wellbeing data, unfortunately, we do not have good time trend data for Mongolia, as

the Gallup World Poll only began polling there in 2005. Trends from 2005 on have been fairly

stable, although with a significant downward dip in 2012. It is also important to note that

wellbeing levels in Mongolia are quite low compared to the rest of the world and even in

comparison to the rest of the transition economies, with Mongolia scoring higher than some

countries such as Bulgaria, Georgia, and Ukraine, but lower than many others, including China,

Kyrgyzstan, and Serbia (see Table 2 below).

Table 2. Best possible life for Mongolia, all available years

Year Obs. Mean Std. Dev. World Rank Transition Countries Rank

2007 0943 4.611 1.690 81/104 21/26

2008 0979 4.392 1.606 98/114 15/16

2010 0995 4.590 1.729 98/125 20/27

2011 0995 5.057 1.672 73/126 16/28

2012 0993 4.785 1.566 98/140 24/30

All years 4,905 4.689 1.668

Source: Gallup World Poll Data, 2008-2009, 2011-2013

Note: “Best possible life” (BPL) measures the respondent’s assessment of her current life relative to her

best possible life on a scale of 0 to 10, where 0 is the worst possible life, and 10 is the best possible life. The

table shows the country means and standard deviations for each year. World rank means that Mongolia

was ranked 81 out of 104 countries in 2007, for example, where Denmark was ranked as being 1 (i.e., having

the highest possible BPL score). Transition countries rank gives the respective rank among transition

countries. Transition countries are defined as in Guriev and Zhuravskaya (2009), and are as follows:

Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic,

Estonia, Georgia, Hungary, Kazakhstan, Kyrgyz Republic, Latvia, Lithuania, FYROM, Moldova, Poland,

Romania, Russia, Serbia and Montenegro, Slovak Republic, Slovenia, Tajikistan, Ukraine, and Uzbekistan.

Given recent world developments, Kosovo and Montenegro are added to this list. Not all countries are

surveyed for all years.

Meanwhile, Mongolia ranks in the top third of countries (41 out of 125) making positive changes

in wellbeing rankings from 2005 to 2015. And compared to other factors such as GDP per capita,

health status, and freedom to make life choices, social support plays an important relative role

in overall determinants of Mongolia’s life satisfaction rankings compared to other countries

(Helliwell, Layard, & Sachs, 2015).

The most recently released Gallup data, for 2013 and 2014, show consistency in Mongolia’s

mean levels of wellbeing – rising slightly to 4.90 in 2013 and then down a little to 4.77 for 2014,

but still in the same range. Mongolia’s average world ranking for 2010-2012 was 102, and then

100 for 2013-2014. That may reflect a slight improvement in ranking, but also the fact that several

additional countries were added to the poll in the last two years (Gallup World Polls data, 2013-

2014).

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 54

3. Data and methods

Our survey was modeled on a wide range of other wellbeing surveys around the world and

included the usual socio-demographic information, as well as an evaluative wellbeing question

(life satisfaction on a 5-point scale) and a hedonic wellbeing question (how happy an individual

felt last week, also on a 5-point scale). Typically, evaluative questions are more susceptible to

question framing effects, and thus the life satisfaction question was asked before the hedonic

(happy last week) question.4 The correlation between the responses to the two questions is .27,

which confirms that while they are indeed related, they are capturing different facets of

wellbeing. For the distribution of responses across these two main wellbeing variables, see Figure

1 below.

Figure 1. Frequency distribution of life satisfaction and “happy last week”

4 For details about question framing and other measurement issues, please see the recent National Academy of Sciences

report on wellbeing measurement, to which one of the authors contributed; Stone and Mackie (2013).

150

300

450

600

750

900

V
e
ry

d
is

a
p

p
o

in
te

(=
1
)

D
is

a
p

p
o

in
te

N
e
it

h
e
r

d
is

a
p

p
o

in
te

d
n

o
r

sa
ti

sf
ie

S
a
ti

sf
ie

V
e
ry

s
a
ti

sf
ie

d
(

=
5
)

C
a
n

't
s

a
y

N
u

m
b

e
r

o
f

re
sp

o
n

d
e

n
ts

"How satisfied are you with your life?"

100

200

300

400

500

600

N
e
v

e
r

(=
1

)

M
a

y
b

e
o

n
ce

O
cc

a
si

o
n

a
ll

M
o

st
d

a
y

E
v

e
ry

d
a

y
(

=
5

)

C
a

n
't

s
a

yN
u

m
b

e
r

o
f

re
sp

o
n

d
e
n

"How happy did you feel last week?"

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 55

While the two questions are related, we used both separately in different regressions (as

discussed below). In the last set of regressions, we also used an additional question about

enjoying life as a control for innate affect/personality traits. In the absence of panel data and the

ability to control for person fixed effects, including a question which gauges positive

affect/optimism in cross section data is a next best approach, which is increasingly common in

the literature.5 For details on the variables in the questionnaire, see the Appendix.

Mongolia has a capital city of Ulaanbaatar and 21 provinces (aimags), which are subdivided

into 329 counties (soums). Soums, in turn, are further divided into bags, which are less formal

administrative units. Our survey was carried out by the Chamber of Commerce and Industry of

Orkhon-Bulgan provinces in Orkhon Province during the period of October-December 2012, and

it is the first-ever conducted survey of its kind in Mongolia. It covered 1,225 randomly selected

respondents between the ages of 15 and 65 from 1,225 households across 20 bags, which

represents 5.1% of all households in the province. Compared to other provinces, Orkhon is

geographically smaller, and is centered around Erdenet, the second largest city in Mongolia.

Hence, Orkhon province is essentially the outer regions of the second largest city in Mongolia

(which is still a relatively small city – 80,000 in population, given Mongolia’s largely rural nature).

Therefore, urban-rural differences, which can be a big issue in some contexts, are less of an issue

in this case, as there are no major differences in urban/rural settings across the province.

Summary statistics of the survey are provided in Table 3 below.

Our baseline regression in Table 4 below is an ordered logistic model, as this specification is

usual for categorical variables which are ordinal but not cardinal in nature. We also re-ran the

same baseline regression with an ordinary least squares (OLS) specification and got essentially

identical results.6 Therefore, we have utilized OLS regressions and both raw and standardized

coefficients throughout the rest of the analyses. The linear specification makes it easier to

compare the coefficients across the equations, as, for example, when we split our sample (Table

5 below). Standardized coefficients, meanwhile, provide a measure of relative influence of the

various explanatory variables. While OLS results usually demonstrate the effect of a one-unit

change in the independent variable on the dependent variable, one cannot interpret the results

the same way when the various independent variables have different scales (as is common with

wellbeing data) and thus it is necessary to standardize the coefficients, following Fields (2004).

5 For a summary of the approach, see Graham and Lora (2009), Chapter 2.
6 It is increasingly common to treat wellbeing data as “cardinal,” at least in practice, if not in theory, and to use OLS

specifications when they are more adequate for the question at hand. See Van Praag and Ferrer-i- Carbonell (2008).

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 56

Table 3: Descriptive statistics

Variable Mean Median Min Max St. Dev N

Measures of Well-Being

Life satisfaction -3.92 4 -1 5 00.69 1,223

Happy last week -3.62 4 -1 5 00.82 1,174

Explanatory Variables

Income -2.91 3 -0 5 01.45 1,226

Relative income -0.02 0 -4 3 01.42 1,226

Unemployed -0.12 0 -0 1 00.32 1,226

Married -0.65 1 -0 1 00.48 1,226

Lives with extended family -0.19 0 -0 1 00.39 1,226

Age 35.81 34 15 65 12.88 1,226

Enjoys life -3.09 3 -1 4 00.53 1,225

Freedom satisfaction -3.62 4 -1 5 00.86 1,209

Dream fulfillment -3.12 3 -1 5 00.95 1,225

Health -3.54 4 -1 5 00.71 1,226

Exercise -1.58 1 -1 5 01.05 1,225

Alcohol use -0.28 0 -0 1 00.45 1,226

Female -0.51 1 -0 1 00.50 1,226

Education -4.54 4 -1 6 01.15 1,226

Home ownership -0.88 1 -0 1 00.32 1,226

Lives in ger dwelling -0.58 1 -0 1 00.49 1,226

Religion -0.69 1 -0 1 00.46 1,226

Income-related stress -0.47 0 -0 1 00.50 1,226

Health-related stress -0.16 0 -0 1 00.36 1,226

Living environment-related

stress
-0.20 0 -0 1 00.40 1,226

Family-related stress -0.08 0 -0 1 00.27 1,226

Job-related stress -0.13 0 -0 1 00.36 1,226

Number of stress triggering

areas
-1.14 1 -0 6 1.06 1,226

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 57

Table 4: Determinants of wellbeing in Mongolia, baseline regressions (ordered logistic and linear)

Ordered Logistic OLS

(1)

Life satisfaction

(2)

Happy last week

(3)

Life satisfaction

(4)

Happy last week

Raw coeff. p-value Raw coeff. p-value Raw coeff. Std. coeff. p-value Raw coeff. Std. coeff. p-value

Income -0.225*** 0.00 -0.127*** 0.00 -0.069*** -0.146*** 0.00 -0.043*** -0.076*** 0.02

Unemployed -0.604*** 0.00 -0.504*** 0.00 -0.198*** -0.093*** 0.00 -0.231*** -0.091*** 0.00

Married -0.541*** 0.00 -0.014*** 0.91 -0.186*** -0.130*** 0.00 -0.011*** -0.006*** 0.85

Lives with extended

family

-0.323*** 0.04 -0.229*** 0.11 -0.095*** -0.054*** 0.04 -0.092*** -0.044*** 0.13

Age -0.055*** 0.08 -0.073** 0.01 -0.019*** -0.358*** 0.05 -0.031*** -0.491*** 0.01

Age2 -0.001*** 0.09 -0.001** 0.03 -0.000*** -0.339*** 0.06 -0.000*** -0.427*** 0.03

Health -0.532*** 0.00 -0.289*** 0.00 -0.166*** -0.171*** 0.00 -0.118*** -0.102*** 0.00

Exercise -0.166*** 0.00 -0.053*** 0.33 -0.050*** -0.077*** 0.01 -0.011*** -0.014**v 0.63

Alcohol use -0.306*** 0.03 -0.088*** 0.50 -0.084*** -0.055*** 0.05 -0.039*** -0.021*** 0.48

Female -0.140*** 0.27 -0.040*** 0.74 -0.039*** -0.028*** 0.32 -0.019*** -0.012*** 0.70

Education -0.027*** 0.64 -0.056*** 0.31 -0.010**v -0.017*** 0.57 -0.028*** -0.039*** 0.23

Lives in ger dwelling -0.184*** 0.16 -0.133*** 0.27 -0.056*** -0.041*** 0.16 -0.052*** -0.032*** 0.30

Home ownership -0.398*** 0.03 -0.177*** 0.31 -0.129*** -0.060*** 0.03 -0.067*v* -0.026*** 0.37

Religion -0.153*** 0.25 -0.001*** 1.00 -0.061*** -0.041*** 0.13 -0.019*** -0.011*** 0.71

Constant -3.189*** 0.00 -3.529*** 0.00

Observations 1,222 1,174 1,222 1,174

R-squared 0.134 0.056

Log likelihood -1124.29 -1348.45

Nagelkerke R-square 0.150 0.069

Cox-Snell R-square 0.129 0.063

* p<0.10, ** p < 0.05, *** p<0.01

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 58

Table 5: Determinants of wellbeing in income subsamples, linear regression

Life satisfaction Happy last week

(1)

< Median income

(2)

> Median income

(3)

< Median income

(4)

> Median income

Raw

coeff.

Std.

coeff.

value

Raw

coeff.

Std.

coeff.

value

Raw

coeff.

Std.

coeff.

value

Raw

coeff.

Std.

coeff.

value

Income -0.082*** -0.072*** 0.13 -0.047*** -0.038*** 0.43 -0.133*** -0.104*** 0.04 -0.035*** -0.021*** 0.68

Unemployed -0.106*** -0.058*** 0.22 -0.296*** -0.114*** 0.02 -0.094*** -0.046*** 0.34 -0.339*** -0.094*** 0.06

Married -0.212*** -0.142*** 0.00 -0.152*** -0.116*** 0.05 -0.005*** 0.003*** 0.95 -0.077*** -0.042*** 0.48

Lives with extended

family

-0.090*** -0.047*** 0.28 -0.090*** -0.059*** 0.22 -0.100*** -0.048*** 0.29 -0.097*** -0.046*** 0.36

Age -0.033*** -0.622*** 0.03 -0.009*** -0.175*** 0.61 -0.056*** -0.935*** 0.00 -0.017*** -0.244*** 0.49

Age2 -0.000*** -0.597*** 0.04 -0.000*** -0.172*** 0.60 -0.001*** -0.859*** 0.00 -0.000*** -0.252*** 0.46

Health -0.147*** -0.146*** 0.00 -0.172*** -0.188*** 0.00 -0.117*** -0.104*** 0.03 -0.133*** -0.104**** 0.04

Exercise -0.069*** -0.094*** 0.04 -0.023*** -0.042*** 0.39 -0.003*** -0.004*** 0.93 -0.034*v* -0.044*** 0.38

Alcohol use -0.116*** -0.068*** 0.13 -0.150*** --0.109*** 0.03 -0.021*** -0.011*** 0.81 -0.133*** -0.070*** 0.17

Female -0.038*** -0.026*** 0.58 -0.042*** -0.035*** 0.48 -0.032*** -0.019*** 0.69 -0.010*** -0.006*** 0.91

Education -0.019*** -0.028*** 0.54 -0.016*v* -0.030*** 0.57 -0.037*** -0.047*** 0.32 -0.006*** -0.008*** 0.89

Lives in ger dwelling -0.071*** -0.044*** 0.33 -0.041*** -0.034*** 0.50 -0.031*** -0.017**v 0.71 -0.143*** -0.083*** 0.11

Home ownership -0.147*** -0.073*** 0.10 -0.211*** -0.090*** 0.06 -0.045*** -0.020*v* 0.65 -0.365*** -0.108*** 0.03

Religion -0.046*** -0.029*v* 0.50 -0.080*** -0.058*** 0.23 -0.001*** -0.001*** 0.99 -0.005*** -0.003*** 0.96

Constant -3.556*** 0.00 --3.024*** 0.00 -3.919*** 0.00 -2.522*** 0.00

Observations 511 430 490 415

R-squared 0.091 0.094 0.075 0.054

* p<0.10, ** p < 0.05, *** p<0.01

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 59

4. Results

4.1 Baseline regression with standard determinants of wellbeing in Mongolia

Our results from the baseline regression in Table 4 above with the standard determinants

demonstrate that the basic determinants of wellbeing are no different in Mongolia than they are

anywhere else, despite the unique context and the dramatic economic and political transition the

country has experienced. We used both ordered logistic and linear regressions and obtained

identical results. With evaluative wellbeing as the dependent variable (which is the most

common specification), income, employment, health (self-reported), and marriage are all

important to wellbeing there, as in other places (Table 4 above). When we look across wellbeing

dimensions, we found, not surprisingly, that the income variable was more significant for

evaluative wellbeing than for hedonic wellbeing. As is noted above, hedonic wellbeing typically

correlates less closely with income (and other agency related variables), as it is more closely

related to day-to-day experience and to innate affect levels than is evaluative wellbeing. Along

these same lines, the coefficient on marriage is positive, and is significant on life satisfaction, but

not significant on “happy last week.” Living with an extended family is not an uncommon

arrangement, and 19% of respondents did indicate that they live in such families. Those living

with an extended family reported lower evaluative wellbeing, but not hedonic wellbeing, in the

baseline regression.

The classic U-shaped age curve, meanwhile, also holds for Mongolia, with the lowest point

in happiness being at 41 years of age. This is on the young end of the curve for most countries (it

is 50 in Russia, for example, 48 years on average for Latin America, and approximately 44 years

of age in the UK and the US), but may in part be explained by fairly low levels of life expectancy

in Mongolia. Life expectancy was 67.34 years in 2012, compared to 69 years in Russia, 73 in China,

and 82 in Japan.7 In addition, given the rapid pace of economic and social changes in recent

decades, strong generational differences exist in personal outlook and values. Indeed, the

standardized coefficients on age are the largest among all the estimated coefficients for both

wellbeing dimensions.

Furthermore, the coefficient on health is positive for both wellbeing dimensions but smaller

in size on “happy last week.” The standardized coefficients show that health is the second most

important determinant of both evaluative and hedonic wellbeing after age. Exercise is positively

correlated with life satisfaction but not with “happy last week,” while alcohol use is negatively

correlated with life satisfaction, but not with “happy last week.”

We also included other variables in the baseline regressions, such as gender, education, home

ownership, dwelling type, and religion. As Table 4 above shows, there was no significant gender

difference in wellbeing. This is a departure from the average for the world as a whole, where

women typically have higher wellbeing levels than men, except in contexts where gender rights

are compromised (Graham & Chattopadhyay, 2013). Mongolia stands out due to its relatively

low degree of gender inequality in certain areas on the one hand, and lack of differences in

wellbeing levels across genders on the other. Due to the nomadic heritage and communist legacy,

Mongolian women actively participate in all arenas of business and society, with the number of

female college graduates, as well as the number of doctors and lawyers exceeding that of men,

as reported by the National Statistical Office of Mongolia.

We also looked at whether living in a ger: the round, portable, felt-covered traditional

dwelling structure (also called a “yurt”) was correlated with wellbeing measures. Most people

7 For the age curve around the world, see Graham (2009). Detailed results on how the age curve was calculated are

available from the authors.

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 60

living in rural areas still reside in this traditional dwelling with little infrastructure, and many

families that have migrated to urban areas have also settled into ger districts on the outskirts of

urban areas. In the baseline regression, there was no significant difference based on living in gers.

In line with evaluative (hedonic) wellbeing typically correlating more (less) closely with agency

related variables, owning one’s home, regardless of whether it is an apartment or a ger is

associated with higher evaluative wellbeing.

Whether an individual reported being associated with a certain religion was not significant,

which is also a departure from the average for the world, in which individuals who associate

with a religion are typically happier than those who do not (except in the context of extreme

religions) (Graham & Crown, 2013). Despite Buddhism being one of the most important

influences on Mongolian culture, with approximately half of the population following Tibetan

Buddhism, and Shamanism/Tengrism experiencing a revival, the ban that was placed on

religious practice under the communist government significantly weakened the role of religion,

and about 40% of the people do not practice any religion, according to various national statistics.

As such, not associating with a religion is as much a norm in Mongolia as associating with one.

Level of education was also not significant (Table 4 above). Despite Mongolia boasting one

of the highest literacy rates in the world at 98%, the education system is still mismatched to the

needs of the economy, as highlighted by both a low-quality and outdated curriculum. This is

common in transition economy contexts, where educational choices made prior to the transition

may not translate into the expected job opportunities post-transition, and hence, education plays

a smaller role in wellbeing than in developed economies. The finding is also in keeping with the

comparative findings across regions, with the coefficient on education displaying less relative

importance in its correlation with wellbeing than in either the EU countries or Latin America

(Graham & Nikolova, 2015).

4.2 Additional determinants of wellbeing in Mongolia

In order to gain more insight into these subjective wellbeing measures, we split our sample into

those respondents above and below median income, since the transition economy context is

clearly different and comparisons can have positive signaling effects, at least for some cohorts,

as Senik (2009) found in Russia (Table 5 above).8 Communist regimes strived to achieve income

equality, and such equality was much emphasized. It may also be that nowadays those above

median income see higher levels of average income as a sign of progress and gains made in the

transition, while poorer respondents may both perceive to be and/or actually be left behind in

the transition process.

There were some notable findings and differences with our split sample specification. We

found that once the sample was split, individual income was mostly not significant, except for

the hedonic wellbeing of those below median income (Table 5 above). It is important to note that

there is little variability in income in the subsamples once we split the sample into above and

below median level of income and exclude the respondents at the median income. Not being

gainfully employed, in contrast, lowered both dimensions of wellbeing only for those above

median income. Marriage remained positive for the life satisfaction of respondents above and

below median income. This contrasts with earlier work we have done on marriage and wellbeing

based on worldwide data, in which we find that the positive coefficient on marriage only holds

8 Since we split the sample into above and below median level of income and exclude the respondents at the median

income, the number of observation varies slightly across subsamples.

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 61

for respondents in wealthier countries and regions and not in poorer ones (Graham &

Chattopadhyay, 2013).

Age (and its square), the variable that explained the most variability in our baseline

regression, was significant for both evaluative and hedonic wellbeing, but only for those below

median income, a finding that is of interest. Health, on the other hand, was significant and

positive for both dimensions of wellbeing for those above and below median income. This is in

keeping with the relative importance of health to wellbeing in virtually all contexts and across

wellbeing dimensions. Exercise was only significant and positive for the life satisfaction of those

below median income, whereas alcohol use was significant and negative for the life satisfaction

of those above median income. While we cannot fully explain these findings, they are suggestive

of different lifestyles, depending on the means that respondents have.

Some recent studies have emphasized the importance of relative income, and utilized the

income rank of a person within his or her community when studying the effect of income on

subjective wellbeing. Given the emphasis on income equality under the communist regime and

the extant wide literature and extensive debate on the relationship between relative income and

wellbeing, we estimate the baseline regression with relative income in Table 6 below as a

robustness test.

Table 6: Determinants of wellbeing with relative income, linear regression

(1)

Life satisfaction

(2)

Happy last week

Raw coeff. Std. coeff. p-value

Raw

coeff.
Std. coeff. p-value

Relative income -0.056*** -0.116*** 0.00 -0.035*** -0.062*** 0.05

Unemployed -0.213*** -0.101*** 0.00 -0.240*** -0.095*** 0.00

Married -0.187*** -0.131*** 0.00 -0.010*** -0.006*** 0.85

Lives with extended

family

-0.092*** -0.053*** 0.05 -0.089*** -0.043*** 0.14

Age -0.019*** -0.356*** 0.05 -0.031*** -0.490*** 0.01

Age2 -0.000*** -0.334*** 0.07 -0.000*** -0.425*** 0.03

Health -0.171*** -0.176*** 0.00 -0.120*** -0.104*** 0.00

Exercise -0.051*** -0.079*** 0.00 -0.011*** -0.015*** 0.62

Alcohol use -0.089*** -0.058*** 0.04 -0.042*** -0.023*** 0.44

Female -0.044*** -0.032*** 0.27 -0.021*** -0.013*** 0.67

Education -0.017*** -0.029*** 0.33 -0.032*** -0.045*** 0.16

Lives in ger

dwelling

-0.106*** -0.076*** 0.01 -0.022*** -0.013*** 0.66

Home ownership -0.144*** -0.067*** 0.01 -0.077*** -0.030*** 0.30

Religion -0.069*** -0.046*** 0.09 -0.014*** -0.008*** 0.79

Constant -3.354*** 0.00 -3.632*** 0.00

Clustered standard

errors

by bag

Yes Yes

Observations 1,222 1,174

R-squared 0.130 0.055

* p<0.10, ** p < 0.05, *** p<0.01

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 62

The effects of inequality on individual welfare – which seem to partially hinge on comparisons

with peers, neighbors, or other relevant cohorts – are rarely captured by large-scale aggregate

measures. In part, this is because inequality signals different things to different people,

depending on the context. There are conflicting results at the country level, with some studies

finding a negative correlation between inequality and life satisfaction, others finding weak

results, and some even finding a positive correlation.9 At more disaggregated regional levels,

income inequality seems to be negatively correlated with life satisfaction in the US, the EU, and

Latin America.10 In contrast, Senik (2004) finds a positive effect of average regional level incomes

in Russia, highlighting the potential role of positive signaling effects in contexts of uncertainty

and transition (which could apply to Mongolia). In another exploration in the transition economy

context, Cojocaru (2012) finds that the wellbeing effects of respondents’ relative rank within

neighborhoods are mediated by their beliefs about whether hard work or connections get one

ahead in life. Those who have faith that hard work leads to upward mobility are not negatively

affected by relative income differences, again likely because of positive signaling effects. More

surprisingly, Clark (2003) also finds a positive correlation between region-level inequality and

life satisfaction in the UK.

Finally, at the neighborhood level, which is what we examine in this paper, there are, again,

different results. Luttmer (2005) finds a negative correlation between average neighborhood level

incomes and life satisfaction in the United States, highlighting the role of negative comparison

effects. Graham and Felton (2006) find that inequality is negatively correlated with life

satisfaction in medium and large cities in Latin America, also suggesting comparison effects, but

positively correlated in the smallest cities, where signaling effects seem to dominate. The work

of Arrow (2009), meanwhile, suggests a related status-seeking channel, whereby people are

willing to forego leisure to “pay” to be above median levels of income.

We compute relative income by subtracting the median community income from a

respondent’s income. Thus the higher the value, the greater the relative difference between the

respondent’s income and the average, and those respondents below median income will have

negative scores. Community here refers to the bag the household belongs to. Bag is the smallest

administrative unit outside of the capital city, and typically there are, at most, a few thousand

individuals in a bag. In our sample, 1,225 respondents come from 20 bags. We replace

respondents’ income with relative income in our baseline model and estimate the regression with

standard errors clustered by bag. It should be noted that income is an ordinal variable on the scale

of 0-5, and thus, there is a limited variability in the relative income variable, with many

respondents’ income being no different from the community median.

The results in Table 6 above display a positive and significant coefficient on relative income

for both dimensions of wellbeing. In other words, having higher levels of income compared to

one’s peers is positively associated with wellbeing, and having incomes below the median is

negatively associated. These findings echo our split sample results and suggest that both

signaling and comparison may be at play, depending on whether respondents are above or below

median levels of income. Not surprisingly, the coefficient was weaker for hedonic wellbeing,

which is typically influenced less by income (and income differentials) than evaluative wellbeing

(Kahneman & Deaton, 2010).

9 See, among others, Alesina et al. (2004); Graham and Felton (2006); Oishi et al. (2011); Schwarze and Harpfer (2007);

and Van Praag and Ferrer-i-Carbonell (2009).
10 See Blanchflower and Oswald (2003); and Graham and Felton (2006), among others.

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 63

Next, we focus on the reported sources of stress in Mongolia. Stress in general is strongly and

negatively correlated with wellbeing, with stress related to circumstances beyond individuals’

control having the worst effects (Haushofer & Fehr, 2014; Stone & Mackie, 2013). Respondents in

our survey were asked whether they had income, health, living environment, family and job-

related stress, and we further collapsed these variables into a reported total number of stress

triggers variable, which we included in our equations. We find, not surprisingly, that the number

of stress triggers or reported stress sources was negatively correlated with both life satisfaction

and “happy last week” in Models 1 and 3 (Table 7 below).

When we look at the specific stress triggers in Models 2 and 4, we find that income, living

environment, and family-related stresses were significant and negative for life satisfaction.

Living environment related stresses include stresses due to the conditions of living environment

and infrastructure, such as inadequate or unsatisfactory electricity and public transportation,

which is common in many developing nations. Especially, in the case of transition countries,

infrastructure was maintained and owned by the state and largely remains so today; quality may

have suffered or failed to improve as part and parcel of the transition.

For hedonic wellbeing (happy last week), only income-related stress was significant and

negative. The direction of causality on the latter is not clear, as some recent work finds that

income matters more to the wellbeing of respondents who are lower on the wellbeing

distribution (Binder & Coad, 2011; Graham & Nikolova, 2015).

Finally, in addition to the usual socio-economic and demographic controls, we also included

a number of variables in our regressions which were intended to capture innate personality traits,

such as optimism and cheerfulness (Table 8 below). These variables were “enjoys life,” “freedom

satisfaction,” and “dream fulfillment.” In the absence of panel data and the ability to include

individual-fixed effects, including an additional question gauged to measure optimism or

pessimism in cross-section data can help control for individual character traits, albeit far from

perfectly. While this is increasingly common in the literature, there is still disagreement among

scholars about whether or not it is appropriate to include perceptions variables in regressions

where the dependent variable is also a subjective variable. We believe that the benefits of

controlling (to the extent we can) for these unobservable traits outweigh the potential risks, and

thus did so in an alternative specification from our baseline regressions.

“Enjoys life” is not a perfect proxy for more commonly used positive affect questions, such

as “smiling or happy yesterday,” but it at least approximates it and is the best fit question

available in the survey.11 “Dream fulfillment” asked whether respondents were able to achieve

their dreams, while “freedom satisfaction” asked whether they were satisfied with their freedom

of expression. These latter two variables capture dimensions of wellbeing – including optimism

and perceptions of agency – which are close to but distinct from life satisfaction.

Not surprisingly, those who enjoy life had higher levels of both evaluative and hedonic

wellbeing, as this variable likely reflects innate personality traits (Table 8 below). Dream

achievement was also positively correlated with both evaluative and hedonic wellbeing. Positive

perceptions tend to correlate together, and dream achievement is a very subjective variable,

reflecting optimism, among other things; causality likely runs in both directions. Freedom of

expression was positively correlated with life satisfaction, but not with “happy last week.”

Freedom of expression captures individuals’ ability to achieve what they want to achieve more

than the quality of their daily lives and/or enjoyment on a day-to-day basis.

11 For a discussion of different interpretations of “enjoys life”, for example, see Steptoe et al. (2012).

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 64

Table 7: Determinants of wellbeing with reported stress triggers, linear regression

Life satisfaction Happy last week

(1)

w/ Number of

stress-causing areas

(2)

w/ Specific

causes of stress

(3)

w/ Number of

stress-causing areas

(4)

w/ Specific

causes of stress

Raw

coeff.

Std.

coeff.

p-value

Raw

coeff.

Std.

coeff.

p-value

Raw

coeff.

Std.

coeff.

p-value

Raw

coeff.

Std.

coeff.

p-value

Income -0.066*** -0.138*** 0.00 -0.065*** -0.137*** 0.00 -0.040*** -0.071*** 0.03 -0.035**v -0.062*** 0.07

Unemployed -0.182*** -0.086*** 0.00 -0.214*** -0.101*** 0.00 -0.222*** -0.087*** 0.00 -0.233*** -0.092*** 0.00

Married -0.191*** -0.133*** 0.00 -0.199*** -0.138*** 0.00 -0.013*** -0.007*** 0.82 -0.030*** -0.018*** 0.59

Lives with extended

family

-0.076*** -0.044*** 0.10 -0.077*** -0.044*** 0.10 -0.076*** -0.037*** 0.20 -0.064*** -0.031*** 0.29

Age -0.013*** -0.253*** 0.17 -0.014*** -0.263*** 0.15 -0.027*** -0.431*** 0.03 -0.026*** -0.414*** 0.04

Age2 -0.000*** -0.225*** 0.21 -0.000*** -0.237*** 0.19 -0.000*** -0.362*** 0.06 -0.000*** -0.355*** 0.07

Health -0.140*** -0.144*** 0.00 -0.145*** -0.149*** 0.00 -0.098*** -0.085*** 0.01 -0.096*** -0.083*** 0.01

Exercise -0.048*** -0.074*** 0.01 -0.049*** -0.075*** 0.01 -0.010*** -0.012*** 0.67 -0.006*** -0.008*** 0.79

Alcohol use -0.077*** -0.050*** 0.07 -0.078*** -0.051*** 0.07 -0.032*** -0.018*** 0.55 -0.045*** -0.025*** 0.41

Female -0.034*** -0.025*** 0.38 -0.033*** -0.024*** 0.40 -0.012*** -0.008*** 0.80 -0.021*** -0.013*** 0.68

Education -0.014*** -0.024*** 0.42 -0.014**v -0.023*** 0.44 -0.031*** -0.044*** 0.17 -0.029*** -0.042*** 0.20

Lives in ger dwelling -0.051*** -0.037*** 0.20 -0.054*** -0.039*** 0.18 -0.057*** -0.035*** 0.26 -0.062*** -0.037*** 0.22

Home ownership -0.117*** -0.055*** 0.04 -0.118*** -0.055*** 0.04 -0.061*** -0.024*** 0.41 -0.062*** -0.024*** 0.40

Religion -0.075*** -0.050*** 0.06 -0.081*** -0.054*** 0.05 -0.008*** -0.005*** 0.87 -0.010*** -0.006*** 0.84

No. of stress triggers -0.099*** -0.152*** 0.00 -0.070*** -0.090*** 0.00

Income-related stress -0.140*** -0.102*** 0.00 -0.197*** -0.120*** 0.00

Health-related stress -0.066*** -0.035*** 0.21 -0.085*** -0.037*** 0.21

Living env.-rel. stress -0.114*** -0.066*** 0.02 -0.072*** -0.035*** 0.24

Family-related stress -0.139*** -0.054*** 0.05 -0.034*** -0.011*** 0.70

Job-related stress -0.055*** -0.027*** 0.33 -0.040*** -0.016*** 0.58

Constant -3.289*** 0.00 3.279*** 0.00 3.597*** 0.00 -3.618*** 0.00

Observations 1,222 1,222 1,174 1,174

R-squared 0.155 0.158 0.063 0.070

* p<0.10, ** p < 0.05, *** p<0.01

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 65

Table 8: Additional determinants of wellbeing in Mongolia, linear regression

(1)

Life satisfaction

(2)

Happy last week

Raw coeff. Std. coeff. p-value Raw coeff. Std. coeff. p-value

Income -0.036*** -0.076*** 0.01 -0.017*** -0.031*** 0.35

Unemployed -0.140*** -0.066*** 0.01 -0.207*** -0.081*** 0.01

Married -0.172*** -0.121*** 0.00 -0.006*** -0.003*** 0.92

Lives with extended family -0.073*** -0.042*** 0.11 -0.057*** -0.028*** 0.34

Age -0.012*** -0.227*** 0.20 -0.026*** -0.409*** 0.04

Age2 -0.000*** -0.180*** 0.30 -0.000*** -0.316*** 0.10

Health -0.112*** -0.116*** 0.00 -0.079*** -0.069*** 0.03

Exercise -0.039*** -0.060*** 0.02 -0.002*** -0.002*** 0.94

Alcohol use -0.067*** -0.045*** 0.10 -0.046*** -0.026*** 0.40

Female -0.048*** -0.036*** 0.19 -0.028*** -0.017*** 0.57

Education -0.001*** -0.001*** 0.97 -0.020*** -0.029*** 0.37

Lives in ger dwelling -0.061*** -0.045*** 0.11 -0.051*** -0.031*** 0.31

Home ownership -0.059*** -0.028*** 0.29 -0.023*** -0.009*** 0.76

Religion -0.026*** -0.017*** 0.51 -0.044*** -0.025*** 0.39

Enjoys life -0.225*** -0.173*** 0.00 -0.163*** -0.106*** 0.00

Freedom satisfaction -0.163*** -0.207*** 0.00 -0.035*** -0.037*** 0.21

Dream fulfillment -0.094*** -0.130*** 0.00 -0.112*** -0.130*** 0.00

Constant -1.963*** 0.00 -2.823*** 0.00

Observations 1,203 1,157

R-squared 0.224 0.080

* p<0.10, ** p < 0.05, *** p<0.01

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 66

It should be noted that democratic government quickly emerged in Mongolia following the fall

of communism, and, as discussed above, Mongolia stands out amongst its neighbors for its high

levels of freedom. Currently, the country is governed by a mixed presidential-parliamentary

system, and despite political crises from time to time, Mongolia is considered free and relatively

stable with little violence.

The standardized coefficients of these additional variables are relatively large, as shown in

Table 8 above, and fit in with the findings discussed above on the relative importance of freedom

in the transition context in general. Our results are essentially unchanged when we re-ran the

regressions without “dream fulfillment” and “freedom satisfaction,” but with “enjoys life”

included in an unreported regression.

5. Conclusions

We built from the burgeoning literature on wellbeing around the world and conducted the first

extensive study of wellbeing in Mongolia, a remote and sparsely populated country that has

experienced an unusually dramatic transition in both its economy and polity in recent decades.

Despite the unique context, we found that the standard determinants of wellbeing were no

different in Mongolia than they are for most countries in the world, with individual income,

health, marital status, and exercise all positively associated with life satisfaction. The same

variables had weaker correlations with “happy last week,” our measure of hedonic wellbeing.

This also accords with previous findings in the literature. The classic U-shaped relationship

between age and happiness also held, with the low point in Mongolia being 41 years, which is

slightly younger than usual, but makes sense, given the lower average level of life expectancy in

Mongolia, and also very significant generational differences due to the rapid economic and social

change in recent decades. Finally, since women typically have higher levels of wellbeing than

men on average (around the world), our finding that they do not have higher levels than men in

Mongolia seems a paradox, given the country’s relatively high levels of gender parity.

We also tested additional contextual variables. When we split our sample into respondents

above and below median levels of incomes, we found some notable differences in our results

from those in other places. We found that relative income differences (each respondent’s

difference from the community median) were positive for those above median income and

negative for those below. The standard interpretation of comparison effects is that they matter

more after people have sufficient income and the “luxury” of worrying about the incomes of

others. Yet the transition economy context is different and comparisons may have positive

signaling effects, at least for those who are doing better than the average, and negative ones for

those who are left behind.

Stress is negative for wellbeing in most contexts, and Mongolia is no exception. Concerns

about income and family stood out as particular triggers of stress, as did concerns about the

living environment. Satisfaction with freedom of expression also surfaced as an important and

positive component of evaluative wellbeing, and may reflect the extent to which Mongolia stands

out for having established democratic governance from the inception of its transition.

In sum, while there are no major surprises in our study, the consistencies that we find in

wellbeing determinants in such a remote and unique setting – which has undergone dramatic

economic and political changes – provide yet another example of how remarkably similar the

correlates of subjective wellbeing are across peoples, cultures, and contexts around the world.

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 67

Acknowledgments

We thank Andrew Clark, Soumya Chattopadhyay, and Milena Nikolova for helpful comments on different

sections of the work.

Authors

Tuugi Chuluun

Loyola University Maryland

Carol Graham

The Brookings Institution

cgraham@brookings.edu

Sarandavaa Myanganbuu

Mongolian National Chamber of Commerce and Industry

Publishing Timeline

Received 22 January 2016

Accepted 22 June 2016

Published 14 October 2016

References

Alesina, A., Di Tella, R., & MacCulloch, R. (2004). Inequality and happiness: Are Europeans and

Americans different? Journal of Public Economics, 88, 2009-2042.

http://dx.doi.org/10.1016/j.jpubeco.2003.07.006

Arrow, K. (2009). Conspicuous consumption, inconspicuous leisure. The Economic Journal, 119

(November), F497-F516. http://dx.doi.org/10.1111/j.1468-0297.2009.02318.x

Binder, M., & Coad, A. (2011). From Average Joe’s happiness to Miserable Jane and Cheerful John: Using

quantile regressions to analyze the full subjective well-being distribution. Journal of Economic Behavior

& Organization, 79(3), 275-290. http://dx.doi.org/10.1016/j.jebo.2011.02.005

Blanchflower, D., & Oswald, A. (2004). Well-being over time in the USA and Britain. Journal of Public

Economics, 88, 1359-1387. http://dx.doi.org/10.1016/S0047-2727(02)00168-8

Clark, A. (2003). Inequality aversion and income mobility: A direct test. Paris, France: DELTA.

Cojocaru, A. (2012). Essays on inequality, social mobility, and redistributive preferences in a transition economy

context (Doctoral dissertation, School of Public Policy, University of Maryland).

De Neve, J. E., & Oswald, A. J. (2012). Estimating the influence of life satisfaction and positive affect on

later income using sibling fixed effects. Proceedings of the National Academy of Sciences, 109(49), 19953-

19958. http://dx.doi.org/10.1073/pnas.1211437109

Fields, G. S. (2004). Regression-based decompositions: A new tool for managerial decision-making. Ithaca, NY:

Department of Economics, Cornell University.

Frey, B., & Stutzer, S. (2002). Happiness and economics. Princeton, NJ: Princeton University Press.

Easterlin, E. (2009). Lost in transition: Life satisfaction on the road to capitalism. Journal of Economic

Behavior and Organization, 71, 130-145. http://dx.doi.org/10.1016/j.jebo.2009.04.003

Graham, C. (2008). The economics of happiness. In S. Durlauf & D. Blume (Eds.), The new Palgrave

dictionary of economics (2nd ed.). London, England: Palgrave-MacMillan.

http://dx.doi.org/10.1057/9780230226203.0702

Graham, C. (2009). Happiness around the world: The paradox of happy peasants and miserable millionaires.

Oxford, England: Oxford University Press.

http://dx.doi.org/10.1093/acprof:osobl/9780199549054.001.0001

Graham, C., & Chattopadhyay, S. (2013). Gender and well-being around the world. International Journal of

Happiness and Development, 1(2), 212. http://dx.doi.org/10.1504/IJHD.2013.055648

mailto:cgraham@brookings.edu
http://dx.doi.org/10.1016/j.jpubeco.2003.07.006
http://dx.doi.org/10.1111/j.1468-0297.2009.02318.x
http://dx.doi.org/10.1016/j.jebo.2011.02.005
http://dx.doi.org/10.1016/S0047-2727%2802%2900168-8
http://dx.doi.org/10.1073/pnas.1211437109
http://dx.doi.org/10.1016/j.jebo.2009.04.003
http://dx.doi.org/10.1057/9780230226203.0702
http://dx.doi.org/10.1093/acprof:osobl/9780199549054.001.0001
http://dx.doi.org/10.1504/IJHD.2013.055648

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 68

Graham, C. & Crown, S. (2014). Religion and wellbeing around the world: Social purpose, social time, or

social insurance? International Journal of Wellbeing, 4(1), 83-99. http://dx.doi.org/10.5502/ijw.v4i1.1

Graham, C., Eggers, A., & Sukhtankar, S. (2004). Does happiness pay? An initial exploration based on

panel data from Russia. Journal of Economic Behavior and Organization, 55, 319-42.

Graham, C., & Felton A. (2006). Inequality and happiness: Insights from Latin America. Journal of

Economic Inequality, 4, 1007-1122. http://dx.doi.org/10.1007/s10888-005-9009-1

Graham, C., & Lora, E. (Eds.). (2009). Paradox and perception: Measuring quality of life in Latin America.

Washington, D.C: The Brookings Institution Press.

Graham, C., & Nikolova, M. (2015). Bentham or Aristotle in the development process? An empirical

investigation of capabilities and subjective well-being. World Development, 68, 163-179.

http://dx.doi.org/10.1016/j.worlddev.2014.11.018

Haushofer, J., & Fehr, E. (2014). On the psychology of poverty. Science, 344(6186), 862-867.

http://dx.doi.org/10.1126/science.1232491

Helliwell, J., Layard, R., & Sachs, J. (2013). World happiness report, 2013. Columbia: Earth Institute.

Helliwell, J., Layard, R., & Sachs, J. (2015). World happiness report, 2015. Columbia: Earth Institute.

Kahneman, D., & Deaton, A. (2010, August). High income improves evaluation of life but not emotional

well-being. Proceedings of the National Academy of Sciences, 107(38), 16489-16493.

http://dx.doi.org/10.1073/pnas.1011492107

Luttmer, E. F. P. (2005). Neighbors as negatives: Relative earnings and well-being, Quarterly Journal of

Economics, 120(3), 963-1002.

Oishi, S., Kesebir, S, & Diener, E. (2011). Income inequality and happiness. Psychological Science, 22(9),

1095-1100. http://dx.doi.org/10.1177/0956797611417262

Schwarze, J., & Harpfer, M. (2007). Are people inequality averse? Evidence from German longitudinal

data on life satisfaction. Journal of Socio-Economics, 36(2), 233-249.

http://dx.doi.org/10.1016/j.socec.2005.11.047

Senik, C. (2004). When information dominates comparison: Learning from Russian subjective panel data.

Journal of Public Economics, 88, 2009-2133. http://dx.doi.org/10.1016/S0047-2727(03)00066-5

Senik, C. (2009). Direct evidence on income comparisons and their welfare effects. Journal of Economic

Behavior & Organization, 72(1), 408-424. http://dx.doi.org/10.1016/j.jebo.2009.04.019

Steptoe, A., Demakakos, P., & de Oliveira, C. (2012). The psychological well-being and functioning of

older people in England. In J. Banks, J. Nazroo, & A. Steptoe (Eds.), The dynamics of aging: Evidence

from the English longitudinal study of aging 2002-10. London, England: IFS.

Stone, A., & Mackie, C. (2013). Subjective well-being: Measuring happiness, suffering, and other dimensions of

experience. Washington, D.C: National Academy of Sciences.

Van Praag, B., Ferrer-i-Carbonell, A. (2008). Happiness quantified: A satisfaction calculus approach. Oxford,

England: Oxford University Press.

Van Praag, B., Ferrer-i-Carbonell, A. (2009). Inequality and happiness. In W. Salverda, B. Nolan, & T.

Smeeding (Eds.), The Oxford handbook of income inequality. Oxford, England: Oxford University Press.

World Economic Forum, (2012). The global gender gap report 2012. Geneva, Switzerland: World Economic

Forum.

http://dx.doi.org/10.5502/ijw.v4i1.1
http://dx.doi.org/10.1007/s10888-005-9009-1
http://dx.doi.org/10.1016/j.worlddev.2014.11.018
http://dx.doi.org/10.1126/science.1232491
http://dx.doi.org/10.1073/pnas.1011492107
http://dx.doi.org/10.1177/0956797611417262
http://dx.doi.org/10.1016/j.socec.2005.11.047
http://dx.doi.org/10.1016/S0047-2727%2803%2900066-5
http://ideas.repec.org/a/eee/jeborg/v72y2009i1p408-424.html
http://ideas.repec.org/s/eee/jeborg.html
http://ideas.repec.org/s/eee/jeborg.html
http://dx.doi.org/10.1016/j.jebo.2009.04.019

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 69

Appendix

Variable Definitions

Variable Definition

Life satisfaction How satisfied are you with your life?

(Very disappointed = 1; Disappointed = 2; Neither disappointed nor

satisfied = 3; Satisfied = 4; Very Satisfied = 5; Can’t say = Missing)

Happy last week Did you feel happy last week?

(Never = 1; Maybe once =2; Occasionally = 3; Most days = 4;

Every day = 5; Can’t say = Missing)

Income Monthly household income

(No income = 0

< 200,000 MNT (≈ < $150) = 1

200,000 – 400,000 MNT (≈ $150-$300) = 2

400,000 – 600,000 MNT (≈ $300-$450) = 3

600,000 – 800,000 MNT (≈ $450-$600) = 4

>800,000 MNT (> $600) = 5)

Average exchange rate of 1,330MNT = 1 USD from 2012 was used to

convert MNT amounts to USD.

Relative income Respondent’s income minus the median income of the bag the

respondent belongs to. Bag is the smallest administrative unit in

Mongolian provinces.

Unemployed Are you unemployed?

(No = 0; Yes = 1)

Married This includes common-law marriages and those living with partners.

(No = 0; Yes = 1)

Lives with extended

family

Whether the respondent lives in an extended family. This includes

three generations living together or living with one’s relatives or in-

laws.

(No = 0; Yes = 1)

Age Respondent’s age, which ranges between 15 and 65.

Enjoys life How much do you enjoy life?

(Not at all = 1; A little bit = 2; Adequate = 3; To the fullest = 4)

Freedom satisfaction How satisfied are you with your ability for free expression?

(Very disappointed = 1; Disappointed = 2; Neither disappointed nor

satisfied = 3; Satisfied = 4; Very Satisfied = 5; Can’t say = Missing)

Dream fulfillment Have you achieved your dreams?

(Haven’t achieved anything = 1; Achieved 25% = 2; Achieved 50% =

3; Achieved 75% = 4; Achieved 100% = 5)

Health How is your health compared to others?

(Very poor = 1; Poor = 2; Okay = 3; Good = 4; Very good = 5)

Exercise On average, how many times do you exercise for more than 30

minutes per week?

(None =1; 1-2 times = 2; 3-4 times = 3; 5-6 times = 4; 7+ times = 5)

Alcohol use Did you use alcohol last month?

(No = 0; Yes = 1)

Wellbeing in Mongolia

Chuluun, Graham, & Myanganbuu

www.internationaljournalofwellbeing.org 70

Variable Definition

Female Respondent’s gender

(No = 0; Yes = 1)

Education Respondent’s educational level

(No education = 1; Primary (1-5 grade) = 2; Middle (5-9 grade) = 3;

Secondary (10-11 grade) = 4; Technical and vocational = 5; Higher

education = 6)

Home ownership Do you own your home?

(No = 0; Yes = 1)

Lives in ger dwelling Does your family live in ger? Ger (or “yurt”) is a traditional round,

portable, felt-covered dwelling.

(No = 0; Yes = 1)

Religion Do you practice a religion?

(No = 0; Yes = 1)

Income-related stress Does not enough income or price increase or cost of children’s

kindergarten and school cause stress in your life?

(No = 0; Yes = 1)

Health-related stress Does family members getting sick and ill or your own poor health

cause stress in your life?

(No = 0; Yes = 1)

Living environment-

related stress

Does home/residence, water, electricity, public transportation, land,

and living environment issues cause stress in your life?

(No = 0; Yes = 1)

Family-related stress Do family arguments and conflicts cause stress in your life?

(No = 0; Yes = 1)

Job-related stress Does too much work or conflict and unpleasant relationships at work

cause stress in your life?

(No = 0; Yes = 1)

Number of stress

triggering areas

The number of areas that caused stress in the respondent’s life,

ranging between 0 and 6. These refer to income, unemployment, job,

health, living environment (e.g., public transportation, electricity)

and family-related stress.