Work and Well-being: A Global Perspective

Abstract

The Global Happiness Policy Report is produced by the Global Happiness Council and contains papers by expert working groups on happiness for good governance. Our chapter on work and well-being provides evidence and policy recommendations on best practices to promote happiness and well-being in the workplace. The first Global Happiness Policy Report was presented at the World Government Summit in Dubai on February 10, 2018.

Full text

This chapter was prepared by Christian Krekel, George Ward, and Jan-Emmanuel
De Neve, and was reviewed by the members of the Workplace Well-being Committee
on the Global Happiness Council. The extensive feedback and comments by the
committee members have much improved the quality of this chapter. We thank the
Gallup Organization for providing access to the Gallup World Poll data set.
De Neve serves as a Research Advisor to Gallup.
75
74
Chapter 5
Work and Well-being:
A Global Perspective
Jan-Emmanuel de Neve
Said Business School, Oxford University
Workplace Well-being Committee
Amy Blankson, Co-founder GoodThink and author of The Future of Happiness:
5 Modern Strategies to Balance Productivity and Well-being in the Digital Era
Professor Andrew Clark, Paris School of Economics
Professor Sir Cary Cooper, Manchester Business School
Dr. James Harter, Chief Scientist of Workplace Management and Well-Being,
Gallup Organization
Dr. Christian Krekel, Post-Doctoral Research Fellow, London School of Economics
Jenn Lim, Advisor at Zappos and co-founder of Delivering Happiness
Dr. Paul Litchfield, Chief Medical Officer at British Telecom and Chair of What
Works Well-being
Ewen McKinnon, Joint head of Cross-cutting Research and Employee Well-being,
UK Cabinet Office
Jennifer Moss, Co-founder Plasticity Labs and author of Unlocking Happiness
At Work
Professor Michael Norton, Harvard Business School
Professor Mariano Rojas, Facultad Latinoamericana de Ciencias Sociales Mexico
George Ward, MIT Institute for Work and Employment Research
Professor Ashley Whillans, Harvard Business School

76
77
1. Introduction
Work and employment play a central role in most
people’s lives. In OECD countries, for example,
people spend around a third of their waking
hours engaging in paid work.1 We not only
spend considerable amounts of our time at work,
employment and workplace quality also rank
among the most important drivers of happiness.
It presents our research on the ways in which
work and workplace quality influence people’s
well-being around the world. It also highlights
a number of best practices that may inspire
policy-makers and business leaders in putting
well-being at the heart of their policies.
Figure 1 illustrates the significance of work: it
reports data from a German survey that asked
people about the importance of different aspects
of their lives for their overall sense of well-being
and satisfaction. 83% of respondents rate work
as either “very important” or “important” for
their well-being, as opposed to 10% and 7%
rating it as less important or even unimportant,
respectively. Further evidence of the significance
of work comes from van Praag et al. (2003), who
use data from the German Socio-Economic
Panel—a nationally representative survey of more
than 11,000 households in Germany that has
been asking respondents about their well-being
since as early as 1984—to study the relative
importance of satisfaction with various life
domains for overall life satisfaction. They find
that the three most important determinants of
life satisfaction are satisfaction with finance (an
area closely related to work), health, and work,
followed by leisure and other life domains.2
Despite the importance of work for people’s
happiness, most do not perceive work as a
particularly enjoyable activity, unfortunately. A
recent study that asked respondents to record
their well-being via a smartphone at random
points in time on a given day found that paid
work is ranked lower than any other of the 39
activities sampled, with the exception of being
sick in bed (Bryson and MacKerron, 2016). In fact,
the worst time of all seems to be when people
are with their boss (Kahneman et al., 2004). Not
surprisingly then, costs of absenteeism and
presenteeism are high: in a recent report for
the UK, it was estimated that absenteeism costs
UK businesses about GBP 29 billion per year,
with the average worker taking 6.6 days off due
to sickness (PwC Research, 2013). Costs of
presenteeism due to, for example, mental health
problems are estimated to be almost twice as
high as those of absenteeism (Sainsbury Centre
for Mental Health, 2007).
What exactly is it about work, and workplace
quality, that makes some jobs less enjoyable
while others more? Answering this question is
not only important because work plays such a
significant role for people’s well-being, but also
because people’s well-being has been found to be
Figure 1: Importance of Work for Wellbeing
(German Socio-Economic Panel, Year 1999)
30 40 5010 200Percent of
Respondents
Very Important
Important
Less Important
Very Unimportant

Global Happiness Policy Report 2018
an important predictor of labor market outcomes
themselves (De Neve and Oswald, 2012), including
job finding and future job prospects when being
out of work (Krause, 2013; Gielen and van Ours,
2014), as well as productivity when being in work
and, ultimately, firm performance (Harter et al.,
2002; Edmans, 2011, 2012; Bockerman and
Ilmakunnas, 2012; Tay and Harter, 2013; Oswald
et al., 2015).3 Being happier also brings with it
objective benefits such as increased health and
longevity, which contribute positively to work
(De Neve et al., 2013; Graham, 2017). Likewise,
well-being has been shown to be positively
associated with intrinsic motivation and creativity
(Amabile, 1996; Amabile and Kramer, 2011; Yuan,
2015). For policy-making, which often boils down
to prioritising attention and resources, it is
important to know which characteristics of work,
and workplace quality, drive people’s well-being,
and should thus be focused upon.
This chapter looks at these characteristics in
a systematic way. We first study the overall
importance of employment itself for self-reported
life evaluation and daily emotions. We then study
how domain-specific measures—job satisfaction
and employee engagement—vary around the
world. Next, in the third and main part of this
chapter, we zoom into workplace quality: here,
we try to find an answer to the question of
exactly which characteristics of work are conducive,
or detrimental, to employees’ well-being. We
conclude by laying out a future research agenda
and putting forward a call for more causal
research on the determinants and benefits of
well-being in the workplace.4
Conclusions are drawn from two datasets, the
Gallup World Poll and the International Social
Survey Program, both of which include the most
important measures of well-being and allow for
international comparisons of working conditions.
Our own analyses are further complemented by
findings from the relevant literature.
2. The Overall Importance
of Employment5
Employment is one of the most important
determinants of our well-being. We can illustrate
this by tabulating the average life evaluation—
measured in terms of the Cantril ladder—for
different employment statuses recorded in the
Gallup World Poll, a survey that is regularly
conducted in more than 160 countries covering
99% of the world’s adult population. The Cantril
ladder asks respondents to imagine themselves
Figure 2a: Importance of Employment Status for Life Evaluation
(Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time)
6
5
4
Life Evaluation
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce

78
79
on a ladder with steps numbered from zero at
the bottom to ten at the top: zero represents the
worst possible life, ten the best.
Figure 2a shows the result of this exercise
for working-age adults: respondents who are
employed and who are working either full-time
for an employer or part-time are most satisfied with
their lives. Respondents who are out of the labor
force are next, but sit clearly below the former two
groups in terms of average life evaluation. In turn,
they are followed by those who are self-employed
full-time and those who are underemployed—
respondents in the latter category work part-time
but would like to work full-time.6 The least happy
are the unemployed: they are almost one whole
life evaluation point below respondents who are
employed and who are working full-time for an
employer—a very large gap.
The devastating effect of unemployment on
people’s well-being is one of the most established
findings in the economic literature on happiness
(see Clark and Oswald (1994) and Winkelmann
and Winkelmann (1998), for example). We know
that life satisfaction does not adapt to being
unemployed (Clark et al., 2008; Clark and
Georgellis, 2013), and that unemployment
leaves a permanent scar even after one regains
employment, in the sense that people who
have been unemployed typically do not return
to the happiness level they had before their
unemployment episode (Clark et al., 2001).7
There are few social norm effects for unemploy-
ment: high unemployment around the unemployed
provides only weak consolation, and does not
become less painful in a social context with high
unemployment (Clark, 2003); for the employed,
it may signal general job insecurity, which in itself
is detrimental to happiness (Luechinger et al.,
2010). Importantly, unemployment is not only a
personal affair: its negative spillovers on other
household members (see Clark (2003), for
example) as well as on society more generally
(see Tay and Harter (2013) or Kunze and Suppa
(2017), for example) are well established.
How does average life evaluation for different
employment statuses differ by gender? As seen in
Figure 2b, women are generally more satisfied with
their lives in every category of employment, and
the relative importance of the different categories
for life evaluation is preserved. A difference,
however, exists for undermployment: women
working part-time but wanting to work full-time
reach about the same happiness level as those
Figure 2b: Importance of Employment Status for Life Evaluation,
by Gender (Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time)
6
5
4
Life Evaluation
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Male Female

Global Happiness Policy Report 2018
being out of the labor force; men, on the other
hand, are less happy when they work part-time but
want to work full-time. Gender norms and lifestyles
may be responsible for such differences.8
Figure 2c plots average life evaluation for different
employment statuses by geographic region.
Needless to say, countries differ greatly in their
political, economic, and cultural institutions, and
aggregate regions may thus be quite heteroge-
neous in terms of countries they include. To
the extent that such differences in institutional
settings pertain to labor markets, for example,
due to differences in active labor market policies
or social safety nets, it may not come as a
surprise that average life evaluation levels differ
for different employment statuses by region. Yet,
with few exceptions, our previous finding holds
across most regions in the world: there exists a
clear-cut importance of being in stable employ-
ment—be it full-time or part-time work—for
people’s well-being over being underemployed,
out of the labor force, or unemployed.
Life evaluation measures such as the Cantril Ladder
make up one element of people’s subjective
well-being. An important further element of
people’s overall happiness is how they experience
their lives day-to-day (Dolan, 2014). The Gallup
World Poll also provides items on positive and
negative affect, constructed from batteries of
yes-no questions that ask respondents about
their emotional experiences the previous day.
For positive affect, these include whether
respondents felt well-rested, whether they were
treated with respect, smiled or laughed a lot,
learned something or did something interesting,
and whether they often felt enjoyment. For
negative affect, these include whether respondents
often experienced physical pain, worries,
sadness, stress, and anger. Indices are then
created by averaging across items, and are
bound between 0 and 100.
Figures 3a to 5a replicate our analyses of life
evaluation for the index of positive affect,
Figures 3b to 5b for that of negative affect.
Turning first to positive affect, Figure 3a, we can
see that the basic insight from our analysis of life
evaluation also holds for how people feel on a
day-to-day basis: respondents who are employed
and who are working full-time for an employer
show the highest positive affect, followed by
Figure 2c: Importance of Employment Status for Life Evaluation, by Region
(Gallup World Poll, Years 2014 to 2016, Weighted by Country; Confidence
Intervals 95%; FT: Full-Time, PT: Part-Time, NZ: New Zealand, CIS:
Commonwealth of Independent States)
8
7
6
5
4
Life Evaluation
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Australia
& NZ
CIS East Asia Europe Latin
America
& Carib
Middle
East &
N Africa
Northern
America
South
Asia
Southeast
Asia
Sub-
Saharan
A fr ic a

80
81
Figure 3a: Importance of Employment Status for Positive Affect
(Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time)
75
70
65
Positive Affect
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Figure 3b: Importance of Employment Status for Negative Affect
(Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time)
40
35
30
25
Negative Affect
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce

Global Happiness Policy Report 2018
Figure 4a: Importance of Employment Status for Positive Affect,
by Gender (Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time)
75
70
65
Positive Affect
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Male Female
Figure 4b: Importance of Employment Status for Negative Affect,
by Gender (Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time)
40
35
30
25
Negative Affect
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Male Female

82
83
Figure 5a: Importance of Employment Status for Positive Affect,
by Region (Gallup World Poll, Years 2014 to 2016, Weighted by Country;
Confidence Intervals 95%; FT: Full-Time, PT: Part-Time, NZ: New Zealand, CIS:
Commonwealth of Independent States)
90
80
70
60
50
Positive Affect
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Australia
& NZ
CIS East Asia Europe Latin
America
& Carib
Middle
East &
N Africa
Northern
America
South
Asia
Southeast
Asia
Sub-
Saharan
A fr ic a
Figure 5b: Importance of Employment Status for Negative Affect, by Region
(Gallup World Poll, Years 2014 to 2016, Weighted by Country; Confidence
Intervals 95%; FT: Full-Time, PT: Part-Time, NZ: New Zealand, CIS: Commonwealth
of Independent States)
8
7
6
5
4
Negative Affect
 Employed FT for Employer
 Employed PT (does not want FT)
 Unemployed
 Employed FT for Self
 Employed PT (wants FT)
 Out of Workforce
Australia
& NZ
CIS East Asia Europe Latin
America
& Carib
Middle
East &
N Africa
Northern
America
South
Asia
Southeast
Asia
Sub-
Saharan
A fr ic a

Global Happiness Policy Report 2018
those who are self-employed full-time and those
who are working part-time, both intentionally
and unintentionally (differences between these
three groups are barely statistically significant at
a conventional level). The lowest positive affect
is again reported by respondents who are
unemployed and who are out of the labor force.
As seen in Figure 3b, a near mirror image is
found for negative affect—the main difference
being that respondents who are unemployed
show the highest negative affect. This “emotional
toll” of unemployment, namely that the
unemployed are sadder than the employed even
when engaging in similar leisure activities, is also
documented in studies using time-use data and
day-reconstruction methods (Knabe et al., 2010;
Krueger and Mueller, 2012). In terms of negative
affect, respondents who are unemployed are
followed by those who are working part-time
but want to work more hours and those who are
out of the labor force.
In line with our findings for life evaluation,
Figures 4a and 4b illustrate that women generally
show more positive and negative affect in every
category of employment; the relative importance
of the different categories for both positive and
negative affect is again preserved. And Figures
5a and 5b illustrate that there are, once again,
large differences in both types of affect across
regions in the world.
So far, we have only looked at descriptive evidence
on the overall importance of employment for
people’s well-being, on average. Needless to say,
average effects may conceal potentially important
effect heterogeneities. More importantly, however,
we cannot make causal statements from
descriptive evidence alone: important observable
characteristics of respondents (for example,
their health) or unobservables (for example,
preferences or personality traits) may explain
both their employment status and their happiness
at the same time. Such omitted characteristics
would inevitably lead to reverse causality and an
overestimation of the true effect of employment
on people’s well-being. Note, however, that our
basic insights continue to hold even if we control
for a rich set of such potentially confounding
characteristics by holding them constant in a
multivariate regression.9 Finally, there is an
established quasi-experimental literature that
exploits plant closures as a source of exogenous
variation to estimate the causal effect of
unemployment on people’s well-being, underlining
its detrimental impact (see Kassenboehmer
and Haisken-DeNew (2008) or Marcus (2013),
for example).
Being in a stable employment relationship, be it
full-time or part-time, provides a sense of purpose
and belonging, social relations, social status, and
a daily structure and routine. This is positively
reflected in how people evaluate their lives
globally, as well as how they feel on a day-to-day
basis. Achieving the desired number of working
hours, for example, by reducing underemploy-
ment, is associated with a well-being premium.
People who are unemployed are worst off: it is
difficult to reconcile this finding with the notion
of voluntary unemployment.
From these basic insights, we can already derive
some important policy implications. In particular,
there is a clear case for active labor market
policies and making job creation a key policy
priority. This could be aided through apprentice-
ship schemes which help younger people to
attain their first job, for example. Potentially
subsidised temporary work schemes could help
the structurally unemployed find their way back
into employment. Temporary work arrangements,
however, should not become entrenched: job
security, as we show below, is an important
predictor of well-being at work. Policies that
would offer (otherwise healthy) firms temporary
financial assistance with the specific aim to avoid
layoffs could be a means to smoothen out
cyclical unemployment in times of economic
crises in order to avoid the heavy psychological
toll on those made redundant as well as to avoid
anxiety for those that remain employed. Such
policies remain to be properly evaluated but
could be found to be highly cost-effective as
they would likely save on unemployment benefits
and on mental health spending.
3. The Global State of Job Satisfaction
and Employee Engagement
We have already seen that average life evaluation
for different employment statuses differs greatly
by region in the world. Different political,
economic, and cultural institutions, especially
those pertaining to the functioning of labor
markets, are most likely driving such differences

84
85
in overall outcomes. It can be expected that, if
we go one step further, we will also find large
differences in how people from different regions
answer questions that are more specific to their
well-being at work.
We are particularly interested in two items that
are sampled in the Gallup World Poll, and that
are more work-specific than overall life evaluation;
these are job satisfaction and employee engage-
ment.10 The former comes from a simple yes-no
question that asks respondents whether they are
“satisfied” as opposed to “dissatisfied” with their
job, while the latter is derived from a set of
formative workplace conditions (such as
opportunity to do what you do best, someone
encouraging your development, and opinions
counting) that are related to a wide range of
business outcomes across organizations. Employee
engagement has three categories: employees
can be “engaged,” “not engaged,” or “actively
disengaged” with their jobs. It is a construct
that goes well beyond job satisfaction: being
engaged with a job requires employees to be
positively absorbed by what they do, and to be
committed to advancing their firm’s interests;
employees who are engaged identify with the
firm and represent it even outside formal
working hours. From a policy perspective, raising
employee engagement therefore represents a
more difficult hurdle to clear than raising job
satisfaction. Needless to say, when looking at
these items, we are confining our analysis to
people who are in work, and who can thus
provide meaningful answers.
Figure 6a shows average job satisfaction levels
by region in the world. We can see that people
who are in work are predominantly satisfied with
their job: the lowest average job satisfaction can
be found in Sub-Saharan Africa; however, even
in this region, about 60% of respondents state
satisfaction as opposed to dissatisfaction with their
job. Sub-Saharan Africa is followed closely by East
Asia (which is dominated by China), South Asia
(which is dominated by India), and Middle East and
North Africa, where average job satisfaction levels
are between 72% and 73%. In the Commonwealth
of Independent States (which is dominated by
Russia) and in Latin America and the Caribbean,
average job satisfaction is slightly higher, at 75%
and 82%, respectively. The front runners are North
America (86%), Europe (86%), and Australia and
New Zealand (87%). Interestingly, these patterns
do not vary significantly when we consider men
and women separately in the analysis.
Figure 6a: Job Satisfaction, by Region
(Gallup World Poll, Years 2010 to 2012, Weighted by Country;
NZ: New Zealand, CIS: Commonwealth of Independent States)
 Satisfied  Dissatisfied
60 80 10020 400Percent of
Working Population
Australia & NZ
CIS
East Asia
Europe
Latin America
& Carib
Middle East & N Africa
Northern America
South Asia
Southeast Asia
Sub-Saharan Africa

Global Happiness Policy Report 2018
Figure 6b replicates Figure 6a for average
employee engagement levels. As noted above,
this indicator is more demanding than job satis-
faction and is a non-binary measure that allows
for increased variation. By and large, the majority
of employees state that they are not engaged
with their job (ranging between 59% and 75%, on
average, depending on region). The regions with
the highest disengagement are East Asia, Europe,
the Middle East and North Africa, and South Asia.
As expected, these regions also count the lowest
shares of engaged employees and the highest
shares of actively disengaged employees. Where
do people fare better? In North America, Latin
America and the Caribbean, and the Common-
wealth of Independent States, about a quarter
of the workforce states engagement with work.
The shares of non-engaged or even actively
disengaged employees are, as expected, compa-
rably low. Again, we find very few systematic
differences when we split the sample by gender.
The seemingly diverging results between job
satisfaction and employee engagement for the
Commonwealth of Independent States highlight
once again that job satisfaction and employee
engagement are very different constructs,
measuring different aspects of well-being at
work. While job satisfaction measures basic
contentment, employee engagement measures
involvement and enthusiasm. The fact that we
find simultaneously high job satisfaction and low
employee engagement levels tells us that, while
most people are content with having a job, a much
lower percentage is emotionally connected with
their work and unlikely to put in discretionary
effort. This also highlights that for a complete
account of well-being in the workplace, a cockpit
of indicators, including additional items such as
purpose or trust rather than a single instrument,
may paint a more nuanced and balanced picture.
Often, however, available data are limited. We
return to this issue in our call for action when
looking ahead at the end of this chapter.
4. Workplace Quality
We have seen the significance of employment in
how people evaluate their lives globally and how
they feel on a day-to-day basis. And we have
seen that there are large differences in these
assessments across regions in the world: not only
does the overall importance of employment for
well-being differ greatly between countries, so
too do satisfaction and engagements levels.
But exactly which job characteristics make certain
jobs less satisfying and others more? To answer
Figure 6b: Employee Engagement Levels, by Region
(Gallup World Poll, Years 2010 to 2012, Weighted by Country;
NZ: New Zealand, CIS: Commonwealth of Independent States)
60 80 10020 400Percent of
Working Population
Australia & NZ
CIS
East Asia
Europe
Latin America
& Carib
Middle East & N Africa
Northern America
South Asia
Southeast Asia
Sub-Saharan Africa
 Actively Disengaged  Not Engaged  Engaged

86
87
this question, we now turn our focus to the work-
place itself and use the latest module on work
orientations of the International Social Survey
Program (ISSP)—a comprehensive, internationally
comparable survey that reports on a wide array
of working conditions alongside well-being for
37 countries across all geographic regions.
Here, we look at job satisfaction as our outcome
of interest. Not only does this measure offer a
distinctively democratic way of asking people
what exactly makes a good job, but it is also
highly correlated with employee retention, an
outcome that is itself highly important to firm
performance. In fact, if we correlate job satisfaction
with the willingness of employees to turn down
a competing job offer, which is also reported in
this survey, we obtain a sizeable correlation
coefficient of about 0.4, suggesting that
employees who are more satisfied with their jobs
are also, to a large extent, more likely to remain
in their jobs. Unlike the previous section, the
ISSP job satisfaction measure is not recorded
by asking employees a simple yes-no question,
but instead offers them more refined answer
possibilities, including “completely satisfied,”
“very satisfied,” “fairly satisfied,” “neither satisfied
nor dissatisfied,” “fairly dissatisfied,” and “very
dissatisfied.” We assign numerical values to these
categories, and use the indicator as a cardinal
measure. To make this measure comparable
across countries, we standardize it such that it
has mean zero and standard deviation one.
Our goal now is to ascertain which specific
elements of workplace quality explain job
satisfaction, our outcome of interest. We set up
a multivariate regression in which we relate job
satisfaction to different domains of workplace
quality as explanatory variables. Building on
Clark (2009), we define 12 of these domains:
1. Pay
2. Working Hours
3. Working Hours Mismatch
4. Work-Life Imbalance
5. Skills Match
6. Job Security
7. Difficulty, Stress, Danger
8. Opportunities for Advancement
9. Independence
10. Interesting Job
11. Interpersonal Relationships
12. Usefulness
At times, a domain includes a single element, as
in the case of working hours (it simply includes
the actual working hours of the respondent), while
at others a domain includes several elements: for
example, Pay includes both the actual income of
the respondent and her subjective assessment
of whether that income is high. In such cases,
we conduct a principle component analysis to
extract a single, latent explanatory factor from
these elements, and then relate job satisfaction
to this factor. In other words, we first establish
which broad domains of workplace quality are
relatively more important for job satisfaction
than others. We then go on to look at the
different elements within these domains in order
to measure their specific contribution to job
satisfaction. We standardize our explanatory
variables such that they have mean zero and
standard deviation one in order to make them
comparable across countries. This also makes
interpretation easier: the coefficient estimate
of an explanatory variable, when squared, now
indicates the variation in job satisfaction that
this variable explains.
To account for potentially confounding individual
characteristics of respondents that may drive
both working conditions and well-being, we
control for a rich set of demographic variables by
holding them constant in our regression. Besides
demographics, differences in job satisfaction
may exist between different occupations and
industries. To be clear, we are not interested in
explaining differences in job satisfaction between,
for example, a manager in the pharmaceutical
industry and a farmer; rather, we are interested in
answering a more fundamental question: which
broad domains of workplace quality are relatively
more important for job satisfaction than others?
(Of course, some of these domains are more
prevalent in certain occupations and industries
than in others). Thus, to isolate the effect of
workplace quality on job satisfaction from any
confounding characteristics, we also control for
occupation and industry. Finally, we further
control for the respective country in which the
respondent lives.11
Before turning to our regression results, we
first look at descriptive evidence that shows the
distribution of job satisfaction and workplace
quality by region in the world.12

Global Happiness Policy Report 2018
As can be seen in Figure 7, there are some
regions that deviate significantly from the
average: Latin America and the Caribbean,
Southeast Asia, the Middle East and Northern
Africa, and Northern America are positive outliers
(differences between these regions are barely
significant at a conventional level); East Asia
(by far) and, to some extent, Australia and
New Zealand are negative ones.
Figures 7a to 7l replicate Figure 7 for the different
domains of workplace quality. As expected,
workplace quality varies greatly across regions in
the world. To get an initial sense of which particular
domains of workplace quality are more strongly
associated with job satisfaction than others, we
pick the most significant outliers from above,
and look into which domains are relatively more
prevalent for them. We take Latin America and
the Caribbean as the positive example and East
Asia as the negative one.
We first look at Latin America and the Caribbean:
the region does not significantly differ from the
average in terms of pay, work-life imbalance, or
independence at work. On the more positive
side, it scores higher in terms of job security,
opportunities for advancement, interestingness
of the job, interpersonal relationships, and
usefulness of work, as well as lower in terms of
working hours mismatch and difficulty, danger,
and stress at work. On the more negative side, it
scores higher in terms of working hours and
lower in terms of skills match.
Interestingly, for East Asia some of these rela-
tionships are reversed. On the positive side, East
Asia scores much higher in terms of pay. On the
negative side, however, it scores higher in terms
of working hours, working hours mismatch,
work-life imbalance, difficulty, stress, and danger
at work, and lower in terms of skills match, job
security, opportunities for advancement, inter-
personal relationships, independence at work,
usefulness, and interestingness of the job.
We now turn to our regression results, and look
more deeply into which of these domains of
workplace quality are relatively more important
for job satisfaction than others. Figure 8 plots
the coefficient estimates obtained from our
regression of job satisfaction on the different
Figure 7: Job Satisfaction, by Region (International Social Survey Program,
Module on Work Orientations, Year 2015, Weighted by Country; Confidence
Intervals 95%; NZ: New Zealand, CIS: Commonwealth of Independent States)
Note: The variable is standardized with mean zero and standard deviation one. The sample is restricted to all
individuals who state that they are working and who report working hours greater than zero.
0.4
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean
 Australia & NZ
 East Asia
 Latin America & Carib
 Northern America
 Southeast Asia
 CIS
 Europe
 Middle East & N Africa
 South Asia
 Sub-Saharan Africa

88
89
Figure 7a: Pay, by Region
0.6
0.4
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean
Figure 7b: Working Hours,
by Region
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean
Figure 7c: Working Hours
Mismatch, by Region
0.2
0.0
-0.2
-0.4
-0.6
Standard Deviation from Mean
Figure 7d: Work-Life Imbalance,
by Region
0.6
0.4
0.2
0.0
-0.2
Standard Deviation from Mean
Figures 7a–7l: Workplace Quality, by Region (International Social Survey Program, Module on Work Orientations, Year 2015,
Weighted by Country; Confidence Intervals 95%; NZ: New Zealand, CIS: Commonwealth of Independent States). See Figure 7
for the Legend. Note: The variables are standardized with mean zero and standard deviation one. Pay, Working Hours Mismatch,
Work-Life Imbalance, Skills Match, Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are
principle components obtained from separate principle component analyses that condense various variables in the respective
domain of workplace quality into a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web
Appendix for summary statistics of the variables. The sample is restricted to all individuals who state that they are working and
who report working hours greater than zero.

Global Happiness Policy Report 2018
Figure 7e: Skills Match
0.2
0.0
-0.2
-0.4
-0.6
Standard Deviation from Mean
Figure 7f: Job Security
0.3
0.2
0.1
0.0
-0.1
-0.2
Standard Deviation from Mean
Figure 7g: Difficulty, Stress,
Danger
0.6
0.4
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean
Figure 7h: Opportunities for
Advancement
0.6
0.4
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean

90
91
Figure 7i: Independence
1.0
0.5
0.0
-0.5
Standard Deviation from Mean
Figure 7j: Interesting Job
0.2
0.0
-0.2
-0.4
-0.6
Standard Deviation from Mean
Figure 7k: Interpersonal
Relationships
0.4
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean
Figure 7l: Usefulness
0.4
0.2
0.0
-0.2
-0.4
Standard Deviation from Mean

Global Happiness Policy Report 2018
domains. The corresponding, more detailed
regression results are available in Table 1 below;
Table 2 employs, instead of the broad domains
of workplace quality, the different constituent
elements within these domains.13
In what follows, we discuss the relative importance
of the different domains of workplace quality for
job satisfaction, including, where appropriate the
different elements within these domains. We look
mostly at their effect on the average employee,
but where interesting, point toward effect
heterogeneities between the employed and the
self-employed (Figure 9a), full-time and part-time
(Figure 9b), and between basic demographic
characteristics such as gender (Figure 9c) and
different levels of education (Figure 9d).
4.1. Pay
It may not come as a surprise that we find pay to
be an important determinant of job satisfaction.
In classic economic theory, labor enters the
utility function negatively, and theory predicts
that individuals are compensated by wages that
equal the marginal product of labor. That said,
pay is not only an important compensation for
the hardship that individuals incur when working
but also an important signal of their productivity.
We thus expect job satisfaction to be higher the
greater the wedge between compensation and
hardship incurred, and the more socially relevant
pay is in a given society.
The importance of pay for job satisfaction seems
universal, with no statistically significant differences
Figure 8: Effect of Workplace Quality on Job Satisfaction
(International Social Survey Program, Module on Work Orientations Year 2015;
Confidence Intervals 95%)
Notes: The figure plots effect estimates obtained from regressing job satisfaction on different domains of workplace
quality. All variables (both left-hand side and right-hand side) are standardized with mean zero and standard
deviation one; regressors are thus beta coefficients. Squaring a regressor yields the respective share in the variation
of job satisfaction that this regressor explains. Pay, Working Hours Mismatch, Work-Life Imbalance, Skills Match,
Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle components
obtained from separate principle component analyses that condense various variables in the respective domain of
workplace quality into a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web
Appendix for summary statistics of the variables. The sample is restricted to all individuals who state that they are
working and who report working hours greater than zero. See Table W3 in the Web Appendix for the corresponding
table with the full set of controls.
0.100.00-0.10
Effect Estimate
Pay
Working Hours
Working Hours Mismatch
Work-Live Imbalance
Skills Match
Job Security
Difficulty, Stress, Danger
Opportunities for Advancement
Independence
Interesting Job
Interpersonal Relationships
Usefulness
0.20 0.30

92
93
between respondents who are employed or
self-employed and working full-time or part-time,
or between gender and different levels of educa-
tion. In our analysis, the domain Pay consists of
two elements: the actual income of respondents
and their subjective assessment of whether that
income is high. Both elements are almost equally
important, but objective income a little more.
Perhaps more surprising is that although pay is
an important determinant of job satisfaction, it is
not the most important one. In fact, it ranks third,
behind interpersonal relationships at work and
having an interesting job. We discuss these
determinants in detail below.
Most people, when asked why they are working,
respond that they are working to earn money.
This is, of course, true, but once working, other
workplace characteristics become more salient, and
thus potentially more important than previously
considered. Experimental research has shown
that intrinsic motivations gain in importance
relative to extrinsic ones (such as income) once
individuals are engaged in an activity (Woolley
and Fishbach, 2015). Particularly, purpose may
be such a characteristic: Ariely et al. (2008)
show, in a laboratory setting, that people who
see purpose in what they do perform relatively
better at work, even in the context of simple,
repetitive effort tasks.14 Using both experimental
and observational data, Hu and Hirsh (2017) find
that employees report minimum acceptable
salaries that are 32% lower for personally mean-
ingful jobs compared to personally meaningless
ones. The important role of purpose may be even
more pronounced when in interplay with good
management practices (Gartenberg et al., 2008),
including employee recognition (Dur et al., 2016).
4.2. Working Hours
As labor enters the utility function negatively,
classic economic theory predicts a negative
relationship between the number of working
hours and well-being. This is precisely what we
find for job satisfaction.
Figure 9a: Effect of Workplace Quality on Job Satisfaction, by Employment
Status (International Social Survey Program, Module on Work Orientations
Year 2015; Confidence Intervals 95%)
Notes: See Figure 8. See Table W5 in the Web Appendix for the corresponding table with the full set of controls.
0.100.00-0.10
Effect Estimate
Pay
Working Hours
Working Hours Mismatch
Work-Live Imbalance
Skills Match
Job Security
Difficulty, Stress, Danger
Opportunities for Advancement
Independence
Interesting Job
Interpersonal Relationships
Usefulness
0.20 0.30
 Employed
 Self-Employed

Global Happiness Policy Report 2018
Interestingly, however, when controlling for all
other domains of workplace quality, the effect
of working hours on job satisfaction is not only
tiny (it ranks as the least important domain of
workplace quality), but statistically insignificant
altogether. This finding is again universal:
there are no statistically significant differences
between respondents who are employed
or self-employed and working full-time or
part-time, or between gender and different
levels of education.
This seems odd at first, but as shown below, is in
line with a growing evidence base that documents
the negative impact of working hours mismatch
and work-life imbalance on well-being.
4.3. Working Hours Mismatch
Rather than the total number of working hours,
what seems to matter more for job satisfaction is
working hours mismatch, defined as the difference
between the actual and the desired number of
working hours.
Individuals differ in their preferences for how
much they want to work, and classic economic
theory assumes that they can freely choose their
desired bundle of labor and leisure hours. Empirical
evidence, however, suggests that this is often not
the case: work contracts, labor market conditions,
and social norms, among others, may affect
choices, and may lead to a realized bundle that
is different from the desired one. In Britain, for
example, more than 40% of employees who work
full-time report a preference of working fewer
hours (Boeheim and Taylor, 2004). In such
situations, theory predicts that individuals end
up on a lower utility level.
We have already seen that employees who
work part-time but prefer to work full-time
evaluate their lives less favourably than those
who intentionally work part-time. We can now
generalize this result and replicate it for job
satisfaction: working hours mismatch has a
significant negative effect on how satisfied
employees are, on average, with their jobs.
Figure 9b: Effect of Workplace Quality on Job Satisfaction, by Working Time
(International Social Survey Program, Module on Work Orientations Year 2015;
Confidence Intervals 95%)
Notes: See Figure 8. See Table W6 in the Web Appendix for the corresponding table with the full set of controls.
0.00-0.20
Effect Estimate
Pay
Working Hours
Working Hours Mismatch
Work-Live Imbalance
Skills Match
Job Security
Difficulty, Stress, Danger
Opportunities for Advancement
Independence
Interesting Job
Interpersonal Relationships
Usefulness
0.20 0.40
 Full-Time
 Part-Time

94
95
It is still unsettled in the literature which is more
detrimental to people’s well-being: underemploy-
ment, as has been found in Germany (Wunder
and Heineck, 2013), or overemployment, as has
been found in Australia (Wooden et al., 2009)
and Britain (Angrave and Charlwood, 2015). In
our analysis, the domain Working Hours Mismatch
consists of two elements: the desire to work
more hours (for more pay) and the desire to
work fewer hours (for less pay). We find that
the latter drives the negative effect of working
hours mismatch on job satisfaction, suggesting
that overemployment is more of an issue than
underemployment. Diverging results in the
literature may point toward the importance of
accounting for differences in institutional set-
tings between countries, including, for example,
differences in labor market regulations (especially
regarding job security), social policy, social
norms, and lifestyles. Note that working hours
mismatch has also been found to have negative
spillovers on other household members (Wunder
and Heineck, 2013).
It turns out that the negative effect of working
hours mismatch on job satisfaction is driven
primarily by the employed over the self-
employed (who probably have more control
over their working hours) and, in line with our
finding for overemployment, by employees
working full-time as opposed to employees
working part-time.
Importantly, there is a gender dimension to
working hours mismatch: its negative effect on
job satisfaction is driven primarily by women.
Evidence shows that women spend considerably
larger amounts of time caring for other house-
hold members (for example, they spend more
than twice as much time on childcare) and doing
routine household work than men, even in cases
where actual working hours are equal between
women and men (OECD, 2014). For women,
achieving a better balance between the actual
and the desired number of working hours would
therefore be an effective means of reducing time
crunches. The fact that working fewer hours
may be detrimental to their long-term career
Figure 9c: Effect of Workplace Quality on Job Satisfaction, by Gender
(International Social Survey Program, Module on Work Orientations Year 2015;
Confidence Intervals 95%)
Notes: See Figure 8. See Table W7 in the Web Appendix for the corresponding table with the full set of controls.
0.100.00-0.10
Effect Estimate
Pay
Working Hours
Working Hours Mismatch
Work-Live Imbalance
Skills Match
Job Security
Difficulty, Stress, Danger
Opportunities for Advancement
Independence
Interesting Job
Interpersonal Relationships
Usefulness
0.20
 Male
 Female
0.30

Global Happiness Policy Report 2018
prospects presents a dilemma, and may—at least
in part—explain the declining life satisfaction of
mothers over the past decades (Stevenson and
Wolfers, 2009).
In sum, we find that working hours mismatch,
in particular overemployment, has a significant
negative effect on job satisfaction. The size of
this effect, however, is rather small: in fact,
working hours mismatch is only ranked 11th out
of the 12 domains of workplace quality in terms
of importance for job satisfaction. If working
hours mismatch is not so bad after all, then
what is? The answer is work-life imbalance, as
discussed below.
4.4. Work-Life Imbalance
Working hours mismatch may not be so
detrimental as long as it does not seriously
interfere with other important domains of life,
especially family. If, however, work and private life
threaten to lose balance, negative consequences
for people’s well-being are large.
Although work-life imbalance ranks only fourth
out of 12 domains of workplace quality in terms
of power to explain variation in job satisfaction,
it is the domain that has the strongest negative
effect on job satisfaction among all negative
workplace characteristics. It is highly significant,
and statistically indistinguishable from exerting
effort in a job that is difficult, stressful, or even
dangerous. The negative effect of work-life
imbalance on job satisfaction seems to be almost
universal: there are no statistically significant
differences between respondents who are
employed or self-employed and between gender.
Perhaps not surprisingly, employees working
full-time are more heavily affected than those
Figure 9d: Effect of Workplace Quality on Job Satisfaction, by Education Level
(International Social Survey Program, Module on Work Orientations Year 2015;
Confidence Intervals 95%)
Notes: See Figure 8. See Table W8 in the Web Appendix for the corresponding table with the full set of controls.
0.00-0.20
Effect Estimate
Pay
Working Hours
Working Hours Mismatch
Work-Live Imbalance
Skills Match
Job Security
Difficulty, Stress, Danger
Opportunities for Advancement
Independence
Interesting Job
Interpersonal Relationships
Usefulness
0.20 0.40
 Low Education
 Medium Education
 High Education

96
97
working part-time, and there is some evidence
that the negative consequences of work-life
imbalance are stronger for workers with low
levels of education.
In our analysis, the domain Work-Life Imbalance
consists of three elements which have a clear
ranking in terms of importance: work interfering
with the family exerts by far the strongest
negative effect on job satisfaction, followed by
the difficulty of taking time off on short notice
when needed. Working on weekends actually
has a positive effect, but is negligible in terms of
effect size.
From our findings on working hours mismatch
and work-life imbalance, we can derive some
important policy implications: policies that target
more supportive and flexible working time
regulations have the potential to considerably
increase people’s well-being. This is especially true
for people who experience disproportionally
more time crunches, including, among others,
women, parents (especially single parents), and
caretakers of other household members such as
elderly. The public policy mix that enables people
to strike a better balance between their work and
private lives can be quite diverse, ranging from
specific labor market regulations on flexible
working times to the provision of infrastructure
such as public transportation in order to reduce
commuting times or early childcare facilities in
sufficient quantity and quality. At the same time,
offering more flexible working times may be a
promising strategy for firms to effectively attract
and retain skilled workers.
Box 1: Work-Life Balance: Is There a Trade-Off Between Flexible Work Practices
and Performance
To answer this question, Bloom et al. (2015)
conducted an experiment at Ctrip, a NASDAQ-
listed Chinese travel agency with more than
16,000 employees. The authors randomly
allocated call center agents who volunteered
to participate in the experiment to work
either from home or in the office for nine
months. They found that working from
home led to a 13% performance increase,
due to fewer breaks and sick days as well
as a quieter and more convenient working
environment. At the same time, job
satisfaction rose and attrition halved.
Conditional on their performance, however,
participants in the experiment were less
likely to get promoted.15 For employees, of
course, this raises the question of whether
flexible work practices are associated with
a career penalty. This does not necessarily
have to be the case: Leslie et al. (2012)
show, in both a field study at a Fortune 500
company and a laboratory experiment, that
flexible work practices result in a career
penalty only if managers attribute their use
as being motivated primarily by reasons
related to personal lives. To the extent that
mangers attribute their use to reasons
related to organizational needs, however,
flexible work practices can actually result
in a career premium. The latter category
includes reasons related to, for example,
work performance and efficiency. Part
of this attribution is communication, and
training supervisors on the value of
demonstrating support for employees’
personal lives while prompting employees
to reconsider when and where to work can
help reduce work-family conflict (Kelly et
al., 2014). Finally, Moen et al. (2011) studied
the turnover effects of switching from
standard time practices to a results-only
working environment at Best Buy, a large
US retailer that implemented the scheme
sequentially in its corporate headquarters:
eight months after implementation,
turnover amongst employees exposed to
the scheme fell by 45.5%. Evidence there-
fore suggests that carefully designed,
implemented, and communicated flexible
work schemes can actually have positive
impacts on organizational performance.

Global Happiness Policy Report 2018
4.5. Skills Match
A job that is asking too much from an employee
can lead to frustration, as can a job that is asking
too little. Matching the demand for and the
supply of skills in a particular job, and enabling
employees to effectively apply the skills they
have or, if necessary, acquire new skills, should
thus be reflected in higher job satisfaction.
This is precisely what we find. Achieving a skills
match in a particular job has a significant positive
effect on how satisfied employees are with that
job. This is again an almost universal finding:
there are no statistically significant differences
between respondents who are employed or
self-employed, between respondents who are
working full-time or part-time, and between
gender. Differences between levels of education
are minor. The domain Skills Match includes two
elements: whether respondents have participated
in a skills training in the previous year and their
subjective assessment of whether their skills
generally match those required in their job. Both
elements matter, but their subjective assessment
a little more.16 Importantly, skills match is not only
directed toward the self but also toward others in
the workplace. In fact, Artz et al. (2017) find that
supervisor technical competence is amongst the
strongest predictors of workers’ job satisfaction.
Willis Towers Watson, a leading human resources
consultancy, estimates that in companies where
leaders and managers are perceived as effective,
72% of employees are highly engaged (Willis
Tower Watson, 2014). On a more abstract level,
the concept of skills match may also be applied
to matching individual character strengths,
although there is as yet little evidence on the
causality of this relationship in organizational
settings.
Although skills match ranks only ninth out of
the 12 domains of workplace quality in terms of
power to explain variation in job satisfaction,
places five to nine are close to each other, and
thus constitute a category of medium importance
for well-being at work.
4.6. Job Security
Slightly more important than skills match is
job security: it ranks sixth out of 12 domains
Box 2: Essential Skills Training: Well-Being Returns and Success Factors
UPSKILL was a workplace literacy and
essential skills training pilot in Canada
(Social Research and Demonstration
Corporation, 2014a). It was implemented as
a randomized controlled trial, involving 88
firms (primarily in the accommodation and
food services sector) and more than 800
workers who were randomly allocated to
receiving 40 hours of literacy and skills
training on site during working hours. The
pilot was not only effective in increasing
basic literacy scores and thus job performance
and retention, but, importantly, also in
increasing mental health: at follow-up,
participants in the treatment group were 25
percentage points more likely than those in
the control group to have reported a signifi-
cant reduction in stress levels. Effects were
particularly pronounced among participants
with low baseline skills. These positive
impacts at the worker level also translated
into positive impacts at the firm level: even
though firms bore the full costs of training
and release time for workers, they incurred a
23% return on investment, primarily though
gains in revenue (customer satisfaction
increased by 30 percentage points), cost
savings from increased productivity (wastage
and errors in both core tasks and adminis-
trative activities were significantly reduced),
and reductions in hiring costs. Besides firms’
commitment to learning and training,
organizations that offered work environments
with high levels of trust gained relatively
more from the program (Social Research
and Demonstration Corporation, 2014b). This
is in line with a growing evidence base on
the importance of trust in the workplace
(Helliwell et al., 2009; Helliwell and Huang,
2012; Helliwell and Wang, 2015).

98
99
of workplace quality, and is thus also part of the
category of medium importance for well-being
at work.
Job security is universally important: we find no
evidence of effect heterogeneities between
respondents who are employed or self-employed
and working full-time or part-time, or between
gender and different levels of education.
The literature shows that the unemployment rate
in a particular region has a significant negative
effect on the life satisfaction of the employed in
that region (Luechinger et al., 2010). This is often
interpreted as a signal of general job insecurity,
which is detrimental to happiness.
4.7. Difficulty, Stress, Danger
Not surprisingly, we find that jobs which are
associated with difficulty, stress, or even danger
are also associated with lower levels of job
satisfaction. This holds true even when controlling
for all other domains of workplace quality,
including pay, working hours, and job security.
This is an interesting finding in and of itself, as
classic economic theory predicts that workers
should be compensated, either monetarily or
non-monetarily, for any job disamenities such
that the net well-being effect is zero. Empirical
evidence on so-called compensating differentials,
however, is rather mixed. In our data, which are
clearly limited, we cannot detect them.
In our analysis, the domain Difficulty, Stress,
Danger consists of two elements: physically
taxing work and stressful work. It turns out that
the latter drives the negative effect of this
domain on job satisfaction; the former, on the
contrary, turns out statistically insignificant. The
fact that stress at work is detrimental to health
is well-established in the literature: for example,
Chandola et al. (2006), in a large-scale prospective
cohort study involving more than 10,000 men
and women aged 35 to 55 who were employed
in 20 London civil service departments, study
the relationship between exposure to stressors
at work and the risk of developing the metabolic
syndrome, a cluster of at least three of five
medical conditions including, among others,
obesity, high blood pressure, and high blood
sugar. They find that employees with chronic
work stress were more than twice as likely to
develop the syndrome 14 years into the study
than those without.
Having a job that is difficult, stressful, or dangerous
ranks fifth out of 12 domains of workplace quality
in terms of power to explain variation in job
satisfaction. It is the domain that has the second
strongest negative effect on job satisfaction
among all negative workplace characteristics, and
ranks directly after work-life imbalance from which
it is—at least in terms of effect size—statistically
not distinguishable. We find little evidence that its
negative impact varies for different people.
4.8. Opportunities for Advancement
We have already seen that being in a stable
employment relationship, be it full-time or
part-time, has positive effects on how people
evaluate their lives globally, as well as how they
feel on a daily basis. Part of why this is the case
is that jobs provide opportunities for advance-
ment, be it steps to climb up the career ladder,
new challenges that give room for personal
development, or others.
Our data do not discriminate between different
types of opportunities for advancement, but
simply ask respondents whether their current
job provides them. This gives respondents the
freedom to interpret the question in whatever
way they themselves find most important.
We find that opportunities for advancement have
a significant positive impact on the average
respondent’s job satisfaction. There is quite some
effect heterogeneity, though: the effect is primarily
driven by respondents who are employed as
opposed to self-employed (probably because
the self-employed are themselves more in
control of which opportunities to create or not)
and by respondents who work full-time as opposed
to part-time. There also seems to be a gradient in
education: opportunities for advancement become
more important for job satisfaction the higher
the level of education. They are, however, equally
important to men and women.
Opportunities for advancement rank seventh out
of the 12 domains of workplace quality in terms of
power to explain variation in job satisfaction.
Perceived progress through well-defined goal-
setting and planning as well as measurable
evaluations—based on clearly defined expectations
and performance—and employee recognition may
increase agency and make the path toward career
advancement more transparent, thereby contrib-
uting positively to well-being at work.

Global Happiness Policy Report 2018
4.9. Independence
Independence at work can have many facets.
Our survey asks respondents to what extent they
can work independently, whether they often
work at home, and whether they have agency
about the organization of their daily work, their
working hours, and their usual working schedule.
We find that independence at work occupies the
middle ground of importance for well-being: it
has a significant positive effect on job satisfaction,
with an effect size similar to skills match, job
security, opportunities for advancement, and
usefulness. It is ranked eighth out of the 12 domains
of workplace quality in terms of power to explain
variation in job satisfaction. Independence at
work seems to be important to everybody:
there are no statistically significant differences
between respondents who are employed or
self-employed and working full-time or part-
time, or between gender and different levels
of education.
In our analysis, the domain Independence includes
eight elements: the subjective assessment of
respondents as to what extent they can work
independently, how often they work at home
during their usual working hours, and whether
the organization of their daily work, their working
hours, and their usual working schedule is
entirely free for them to decide as opposed to
fixed. Some of these elements are important
while others are not. There also seems to be a
ranking of importance: we find that the positive
effect of independence at work on job satisfaction
is driven primarily by whether respondents
report that they can freely organize their daily
work, followed by their subjective assessment
as to what extent they can work independently.
The nature of having discretion about the usual
working schedule is more complex: we find that
both full discretion and no discretion at all have
a negative impact on job satisfaction. Here, it
seems that the reference category—having
limited discretion—yields a higher job satisfaction
than both ends of the spectrum.
Independence at work is related to the concept
of job crafting (Wrzesniewski and Dutton, 2001),
and the question of whether organizations
Box 3: Does Autonomy over Working Schedules Raise Employee Well-Being?
STAR (“Support. Transform. Achieve.
Results”) was a flexible working practices
pilot developed by the interdisciplinary
Work Family and Health Network (King et
al., 2012). It aimed to (i) increase employees’
control over their working schedule, (ii) raise
employee perceptions of supervisor support
for their personal and family lives, and
(iii) reorient the working culture from face
time to results only. Eight hours of preparatory
sessions encouraged managers and their
teams to identify new, flexible work practices,
for example, by communicating via instant
messenger or planning ahead for periods of
peak-demand more effectively. The pilot
was implemented as a group-randomized
controlled trial in a Fortune 500 company,
involving 867 IT workers who were, including
their entire team, allocated to either the
intervention or business-as-usual and
followed for over a year. Moen et al. (2016)
find that the intervention significantly
reduced burnout by about 44% of a
standard deviation while raising job
satisfaction by about 30%. These large
effect sizes were partially mediated by
decreases in family-to-work conflict and,
perhaps less surprisingly, increases in schedule
control. There is also some evidence that
the intervention decreased perceived stress
and psychological distress. Although it has
not been evaluated with respect to employee
performance (possibly because it is difficult
to measure performance in the given
context), recent experimental evidence
(see Bloom et al. (2015), for example)
suggests that, in a very similar context,
giving employees more autonomy over
where and when to work can have strong,
positive performance impacts.

100
101
should give their employees a certain freedom to
design their jobs based on personal needs and
resources. Studies have shown that enabling
employees to craft their jobs in this way can have
positive benefits in terms of increased employee
engagement and job satisfaction as well as
decreased likelihood of burnout (Tims et al.,
2013). More generally, the concept of individual
job crafting may be transferred to the level of the
entire organization, in the sense of organizational
design. It can also be applied to the physical
environment: Knight and Haslam (2010) studied,
in an experiment involving different office spaces,
the effect of giving employees the opportunity
to design their physical working environment. In
line with the notion of social identity, they found
that employees who were randomly allocated to
the crafting condition showed higher organizational
identification, job satisfaction, and productivity,
measured in terms of task performance. Indepen-
dence at work has also been identified as a
contributing factor to creativity (Amabile et al.,
1996). Evidence is thus rather positive about
independence; its precise impact, however, is
probably highly context-specific.
4.10. Interesting Job
It should not come as a big surprise that having
an interesting job is positively associated with
being more satisfied with it.
But it is astonishing just how important interest-
ingness is. Amongst all positive workplace
characteristics, it has the second strongest effect
on job satisfaction, right after interpersonal
relationships at work (from which it is, in terms
of effect size, not statistically distinguishable),
and thus ranks second out of the 12 domains of
workplace quality in terms of power to explain
variation in job satisfaction. There is little evidence
that the impact of interestingness varies for
different people: having an interesting job is
important to everybody.
Note that interestingness is not the same as
purposefulness. A job can score both high on
being interesting and low on being purposeful.
In contrast to interestingness, purposefulness is
best described in terms of a long-term alignment
between a job and an individual’s own evolutionary
purpose in the sense of doing something greater
than self.
4.11. Interpersonal Relationships
In most jobs, employees interact, in one way or
another, with supervisors, co-workers, or clients.17
The way in which these interactions occur, and
interpersonal relationships are maintained,
proves to be the most important determinant of
employee job satisfaction.
Interpersonal relationships have a sizeable and
significant positive effect on the job satisfaction
of the average employee. They rank first out of
our 12 domains of workplace quality in terms of
power to explain variation in job satisfaction. The
size of the effect, however, is statistically not
different from that of having an interesting job,
which ranks second. Interpersonal relationships
are particularly important for the employed as
opposed to the self-employed (probably because
the self-employed can, if necessary, avoid interac-
tions) and employees who are working full-time
as opposed to those who are working part-time
(probably because people become relatively
more important the more time is spent with
them). There is no gender dimension to interper-
sonal relationships—they are equally important to
men and women—nor does their importance for
job satisfaction vary by educational level.
In our analysis, the domain Interpersonal
Relationships consists of three elements: contact
with other people in general, the respondents’
subjective assessment of their relationship with
the management, and the equivalent subjective
assessment of their relationship with co-workers.
The driver behind the positive effect of interper-
sonal relationships on job satisfaction is, by
far, the relationship with the management; the
relationship with co-workers is, although
important, only half as important. This is in line
with evidence showing that about 50% of US
adults who have left their job did so in order to
get away from their manager (Gallup News,
2015). Contact with other people in general
seems to matter less for job satisfaction.
4.12. Usefulness
How important is pro-sociality—doing something
that is beneficial for other people or for society
at large—when it comes to job satisfaction?
Pro-social behavior is behavior intended to
benefit one or more individuals other than
oneself (Eisenberg et al., 2013). This type of

Global Happiness Policy Report 2018
behavior can cover a broad range of actions such
as helping, sharing, and other forms of cooperation
(Batson and Powell, 2003).18 It has been shown to
have positive well-being benefits at the individual
level (Meier and Stutzer, 2008). At the societal
level, it can help build social capital through
fostering cooperation and trust, and social
capital is linked to higher levels of well-being in
societies (Helliwell et al., 2016, 2017). Pro-sociality
is not the same as purpose (although they overlap
to a very large extent): whereas pro-sociality is
always directed toward others, purpose could, in
the narrower sense, only be directed toward the
self. That said, a job can score both high on
individual purpose and low on pro-sociality. In
reality, however, most jobs probably score either
high or low on both constructs.
Box 4: How the Relationship Between Managers and Employees Affects
Well-Being at Work
Managers can have many functions: for
employees, they may provide training,
advice, and motivation (Lazear et al., 2015).
To effectively fulfill these functions, managers
should be competent. Artz et al. (2017)
study the relationship between managers’
technical competence and employees’ job
satisfaction using the Working in Britain
Survey in the UK and the National Longitudinal
Study of Youth in the US. They find that a
manager’s technical competence—measured
in terms of whether the manager worked
herself up the ranks, knows her job, or could
even do the employee’s job—is the single
strongest predictor of an employee’s job
satisfaction. In terms of effect size, having
a competent boss is even more important
for job satisfaction than having friendly
colleagues. In a study on the National Health
Service in England, Ogbonnaya and Daniels
(2017) find that trusts (the organizational
entities that make up the National Health
Service) which make the most use of people
management practices are over twice as
likely to have staff with the highest levels of
job satisfaction as compared to those which
make the least use of these practices. People
management practices refer to training,
performance appraisals, team working, clear
definition of roles and responsibilities,
provision of autonomy in own decision-
making, and supportive management that
involves staff in organizational decisions.
Importantly, they are also three times more
likely to have the lowest levels of sickness
absence, and four times more likely to have
the most satisfied patients. White and
Bryson (2013) confirm this finding for a
wider range of organizations in Britain, using
an index constructed from various domains of
human resource management—participation,
team working, development, selection, and
incentives—and nationally representative,
linked employee-employee data: firms with
more human resource practices in place tend
to score higher in terms of employees’ job
satisfaction and organizational commitment
(although the relationship seems to be
non-linear). Fairness and transparency in
managerial decision-making seems to be an
important factor: Heinz et al. (2017) conduct
a field experiment in which the authors set
up a call center to study the impact of
treating some employees unfairly on the
productivity of others. They set up two work
shifts, and randomly lay off 20% of employees
between shift one and two due to stated
cost reductions (which, as confirmed by
interviews with actual HR managers, is
perceived as unfair). The productivity of the
remaining, unaffected workers, which are
notified by this decision at the beginning of
the second shift, drops by about 12 percent.
The effect size of the productivity decline is
close to the upper bound of the direct
effects of wage cuts.

102
103
We can replicate this finding for well-being at
work: doing something that is beneficial for
other people or for society at large is associated
with higher levels of job satisfaction, on average.
However, in line with the notion of humans as
conditional co-operators (Fehr and Fischbacher,
2003), the size of this effect is rather small.
Usefulness ranks only tenth out of our 12 domains
of workplace quality in terms of power to explain
variation in job satisfaction. There is also quite
some effect heterogeneity: doing something
useful is more important for the job satisfaction
of the employed as opposed to the self-employed
(probably because the self-employed have, in the
first place, more choice over which activities to
engage in or not) and employees who are
working full-time as opposed to those working
part-time. Pro-sociality also becomes more
important the higher the level of education.
There are, however, no significant differences
between gender.
In our analysis, the domain Usefulness consists
of two elements: helping other people and being
useful to society. Both are important, but being
useful to society slightly more so.
There is growing literature on pro-sociality in the
workplace. Anik et al. (2013) studied the impact
of pro-social bonuses—a novel type of bonus
spent on others rather than one-self—on well-be-
ing and performance. In a field experiment at a
large Australian bank, the authors found that
employees who were randomly allocated to
receive bonuses in the form of (relatively small)
financial donations to be made to local charities
showed significant, immediate improvements
in job satisfaction and happiness compared to
employees who were not given these bonuses. In
two follow-up experiments, one involving sports
teams in Canada and another involving a sales
team at a large pharmaceutical company in
Belgium, they found that spending bonuses on
team members rather than oneself led to better
team performance in the longer term. The
finding that spending money on others can buy
you happiness has also been shown by Dunn et
al. (2008): the authors find that pro-social
spending in the form of gifts to others or finan-
cial donations to charities is positively correlated
with general happiness; longitudinally, (arguably
otherwise comparable) employees who
unexpectedly received a profit- sharing bonus
and spent more of it pro-socially experienced an
increase in general happiness, even after con-
trolling for income and the amount of the bonus.
Two other intervention studies stand out:
Gilchrist et al. (2014) studied the impact of pay
raises—masked as gifts—on performance in a
setting where there were no future employment
possibilities. The authors hired one-time data
entry assistants on an online platform for free-
lancers, and then randomly allocated them into
different experimental conditions, one involving
an unexpected, benevolent pay raise. They found
that freelancers allocated to this condition
entered 20% more data than those who were
either initially offered the same pay or initially
offered a lower pay, both of which performed
equally. In other words, simply paying more at
the outset did not elicit higher task performance,
but an unexpected pay raise masked as a benev-
olent gift did. Grant (2008), in a randomised field
experiment involving fundraisers at a university,
found that bringing together fundraisers and
beneficiaries to show the former the purpose of
their work significantly increased their subsequent
task performance.
How organizations can organize work to make
it more fulfilling and connect people with the
pro-social impact they may have, for example, by
providing incentives to elicit behaviors that help
accumulate altruistic capital (Ashraf and Bandiera,
2017), is a promising area of research.

Global Happiness Policy Report 2018
Table 1: Effect of Workplace Quality on Job Satisfaction, Aggregated Domains
Workplace Quality Effect on
Job Satisfaction
Ranking of Importance for
Job Satisfaction
Pay 0.131*** 3
(0.0161)
Working Hours -0.0107 12
(0.0104)
Working Hours Mismatch -0.0271** 11
(0.0106)
Work-Life Imbalance -0.106*** 4
(0.00681)
Skills Match 0.0474*** 9
(0.00880)
Job Security 0.0734*** 6
(0.00906)
Difficulty, Stress, Danger -0.0918*** 5
(0.0105)
Opportunities for Advancement 0.0598*** 7
(0.0119)
Independence 0.0551*** 8
(0.0109)
Interesting Job 0.267*** 2
(0.0231)
Interpersonal Relationships 0.281*** 1
(0.0145)
Usefulness 0.0399*** 10
(0.0103)
Constant Yes
Controls Yes
Occupation Fixed Effects Yes
Industry Fixed Effects Yes
Country Fixed Effects Ye s
Observations 16,326
Adjusted R-Squared 0.422
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation
one; regressors are thus beta coefficients. Squaring a regressor yields the respective share in the variation of job
satisfaction that this regressor explains. Pay, Working Hours Mismatch, Work-Life Imbalance, Skills Match, Difficulty,
Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle components obtained from
separate principle component analyses that condense various variables in the respective domain of workplace quality
into a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web Appendix for
summary statistics of the variables. The sample is restricted to all individuals who state that they are working and
who report working hours greater than zero. See Table W3 in the Web Appendix for the full set of controls.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

104
105
Table 2: Effect of Workplace Quality on Job Satisfaction, Disaggregated Domains
Workplace Quality Effect on
Job Satisfaction
Pay
High Income 0.0866***
(0.0122)
Individual Income
(Natural Log)
0.105**
(0.0506)
Working Hours
Working Hours
(Natural Log)
-0.0105
(0.00980)
Working Hours Mismatch
Wants to Work Same Hours Reference
Category
Wants to Work More Hours -0.00979
(0.00697)
Wants to Work Less Hours -0.0297***
(0.00996)
Work-Life Imbalance
Working on Weekends 0.0169**
(0.00699)
Work Interfering With Family -0.109***
(0.00935)
Difficulty of Taking Time Off -0.0385***
(0.00900)
Skills Match
Skills Match 0.0476***
(0.00920)
Skills Training 0.0190**
(0.00878)
Job Security 0.0700***
(0.00847)
Difficulty, Stress, Danger
Hard Physical Work -0.00739
(0.0119)
Stressful Work -0.0853***
(0.0113)
Opportunities for Advancement
Opportunities for Advancement 0.0538***
(0.0114)
Independence
Independent Work 0.0275**
(0.0106)
Working From Home -0.00996
(0.0105)
Daily Work Flexible Reference
Category
Daily Work Fixed -0.0112
(0.00846)
Daily Work Free 0.0386***
(0.0100)
Working Hours Flexible Reference
Category
Working Hours Fixed -0.00195
(0.00742)
Working Hours Free -0.00270
(0.00835)
Working Schedule Flexible Reference
Category
Working Schedule Fixed -0.0212**
(0.00798)
Working Schedule Free -0.0167**
(0.00793)
Interesting Job
Interesting Job 0.265***
(0.0221)
Interpersonal Relationships
Contact With Other People 0.00561
(0.00891)
Relationship With Management 0.222***
(0.0114)
Relationship With Co-Workers 0.0906***
(0.0116)
Usefulness
Helping Other People 0.0256***
(0.00901)
Being Useful to Society 0.0359***
(0.00853)
Constant Yes
Controls Yes
Occupation Fixed Effects Yes
Industry Fixed Effects Yes
Country Fixed Effects Yes
Observations 16,326
Adjusted R-Squared 0.438
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand ide and right-hand side) are standardized with mean zero and standard deviation
one; regressors are thus beta coefficients. Squaring a regressor yields the respective share in the variation of job
satisfaction that this regressor explains. The sample is restricted to all individuals who state that they are working and
who report working hours greater than zero. See Table W4 in the Web Appendix for the full set of controls.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

Global Happiness Policy Report 2018
5. Looking Ahead
Studying well-being at work is important, not
only because work and workplace quality play
such a significant role for people’s well-being,
but also because people’s well-being is an
important predictor of outcomes related to
worker productivity and firm performance.
Harter et al. (2010), using a large longitudinal
dataset that includes 141,900 respondents within
2,178 business units of ten large organizations
across industries, study the relationship between
perceived working conditions of employees and
firm-level outcomes. They find that working
conditions—including overall satisfaction within
the organization—are predictive of key outcomes
such as employee retention and customer loyalty.
Importantly, Harter et al. (2010) are able to show
that the effect tends to run from working condi-
tions to firm-level outcomes rather than the
other way around—this is suggestive of a causal
impact. The strength of the relationship is not
trivial: in a previous meta-analysis, Harter et al.
(2002) estimate that business units in the top
quartile on employee engagement conditions
realize between one and four percentage points
higher profits and between 25% to 50% lower
turnover than those in the bottom quartile.
These findings have direct implications for
managerial practice: Frey (2017) argues that
managers should create workplaces that are
conducive to well-being, for example, by supporting
workers’ independence and creativity or by
fostering interpersonal relationships at work. At
the same time, work should not be so demanding
and burdensome that workers are unable to
enjoy their leisure time; providing more flexible
working hours may be a means to striking a
better balance between work and life. Income
provided should be sufficient to lead a good life
with respect to material standards. All of these
factors have been found to be conducive to
well-being at work, although to varying degrees,
as presented here and reviewed elsewhere (see
OECD (2017b), for example). At the same time,
however, Frey (2017) argues that managers
should not engage in directly trying to maximize
the happiness of stakeholders (which can be
subject to manipulation); rather, they should lay
the foundations within organizations for stake-
holders to achieve happiness in the way they
themselves choose. The importance of autonomy
therefore applies to the question of how to
achieve happiness itself.
The importance of work, and workplace quality,
for well-being and, in turn, the importance of
well-being for individual-level labor market
outcomes as well as key firm-level outcomes
makes a cautious case for active public policy
intervention. Independent staff well-being audits
may be a means to raising awareness for well-be-
ing at work. Awards for work environments that
are conducive to well-being may also be bestowed
on single managers or entire organizations
(Gallus and Frey, 2016; Frey and Gallus, 2017).
Systematic measurement of well-being within
organizations may serve as a diagnostic tool, for
example, to uncover well-being inequalities within
organizations, which have been found to be a
powerful driver of behavior at the community
level and may be relevant to organizations just
as well. It may also serve as a vehicle to pave
the way towards interventions, directed at one or
more domains of workplace quality. The evidence
presented here and reviewed elsewhere (see
Arends et al. (2017) or OECD (2017b), for example)
suggests that workplace quality has rather
positive impacts on productivity and performance,
in line with recent experimental evidence in
various contexts (Bloom et al., 2015; Oswald et al.,
2015). Ultimately, however, more experimental
evidence from the field is needed in order to be
able to make strong causal claims about the
relationship between workplace quality, well-being,
and its objective benefits for both individuals
and firms. In next year’s chapter on work and
well-being for the Global Happiness Policy
Report, we aim to look more closely into these
objective benefits, in order to evaluate and
motivate the economic case for placing
well-being at the core of business practices.
This chapter can only offer a cautious exploration
into the nexus between work and well-being.
Clearly, there are methodological issues: first,
and foremost, the evidence presented here is
mostly descriptive, and from descriptive evidence
alone we cannot make causal statements. There
may be observable characteristics of respondents
or, more importantly, unobservable characteristics
that explain both their work status and their
well-being at the same time. Such omitted
characteristics would inevitably lead to reverse
causality. We need longitudinal data—repeated
observations of the same individuals over time—

106
107
to get closer to causal effects, and ideally, some
sort of randomized experimental intervention or
policy change as an exogenous variation in order
to reduce concerns about self-selection and
omitted variables. We bypassed this issue by
presenting, where available, supporting evidence
from causal-design studies in the literature.
Our tools are also limited in other dimensions.
Not only are available datasets typically limited
in terms of types of outcomes they offer (most
datasets do not include simultaneous evaluative,
experiential, and eudemonic measures of
well-being), but also in terms of country coverage
(a distinctively Western focus). The latest module
on work orientations of the International Social
Survey Program, which we used to study the
effect of workplace quality on well-being, includes
only job satisfaction as a domain-specific,
evaluative measure of well-being. It is quite
possible, however, that some workplace qualities
are more likely to strongly impact eudemonic
measures of well-being such as purpose. We
cannot verify this with our data, and importantly,
cannot check which construct is relatively more
important for which domain of workplace quality.
Ultimately, we need a cockpit of standardized
measures of evaluative, experiential, and
eudemonic measures—like the ONS-4—to lend a
more complete picture of well-being at work.19
In terms of country coverage, the latest module
on work orientations of the International Social
Survey Program is clearly limited: for example,
the only country included in the Sub-Saharan
Africa region is South Africa. Obviously, this
gives a biased picture of work-place quality in
the region. Further, the informal sector, which
by far comprises the largest part of the labor
market in many least developed countries is
completely ignored. Concerning variables on
workplace quality, most datasets today focus
on rather standard items, ignoring more modern
elements of labor markets related to technology
and the future of work such as aspects pertaining
to the so-called “gig economy” or (fear of)
automation and artificial intelligence. Items
sampled in different surveys are also quite
heterogeneous. The OECD Guidelines on Measuring
the Quality of the Working Environment are
therefore a right step toward establishing a
unified framework for measuring workplace
quality, focusing on objective job attributes and
outcomes measured at the individual level
(OECD, 2017b). These guidelines divide job
characteristics into six broad categories, including
the physical and social environment of work,
job tasks, organizational characteristics,
working-time arrangements, job prospects, and
intrinsic job aspects.
Finally, questions remain regarding external
validity: while there are few datasets that are as
comprehensive as the International Social Survey
Program, country-score comparisons with other
datasets show that some of its items have low
convergent validity. Note, however, that similar
findings on the relationship between workplace
quality and job satisfaction have been identified
by De Neve and Ward (2017) using the European
Social Survey. Future research should be directed
toward identifying similar patterns in other
datasets. Importantly, this research should be
seen as an ongoing endeavour: the composition
of the labor supply changes continuously, for
example, as more and more millennials with
preferences different from previous generations
enter the labor force.
In view of these limitations, we end this report
by looking ahead, and putting out a call for
action: we call upon people in academia,
business, and government to work together in
expanding the causal evidence base on work
and well-being. Academics and businesses, for
example, could cooperate and test how
modifications to work processes and practices
affect worker well-being, and ultimately,
performance. Candidates for such modifications
should be guided by theory, and tested in such
a way as to be subject to rigorous impact
evaluation through randomized controlled trials.
This way, we can avoid issues of omitted
characteristics and self-selection, and identify
causal effects of work and workplace quality on
well-being and performance. It will be important
to establish a common set of measures, covering
evaluative, experiential, and eudemonic
measures of well-being, to be used across impact
evaluations of trials. And it will be important to
record and report the costs of these trials (less
the costs of impact-evaluating them). This will
allow for benchmarking interventions in terms
of cost-effectiveness, and rank interventions
according to those which buy more worker
well-being and performance per dollar invested.
Evidence from behavioral science suggests that
seemingly small, low-cost (or even costless)

Global Happiness Policy Report 2018
changes in daily work routines could produce
large gains in well-being and performance.
Partly, our vision is already reflected in academic
practice: in business schools throughout the
world, experimental methods make their way
into curricula, as is the case with A/B testing in
marketing, for example. Knowledge generated
by way of such trials should be shared openly
as best practices, and doing so should be
incentivised. Governments can also become
active players by introducing well-being
interventions within the civil service, which
could also help to promote happiness more
widely in society. After all, a happy and engaged
civil service is an obvious starting point for
being able to deliver on policies that aim to
put well-being at the heart of policy-making.

108
109
Endnotes
1 See OECD (2017a) for data on daily time use in OECD
countries.
2 See Table W1 in the Web Appendix for the respective table
from van Praag et al. (2003).
3 See Tenney et al. (2016) for a review on the relationship
between people’s wellbeing and labor market outcomes, as
well as Judge et al. (2001) and Harrison et al. (2006) for
recent meta-analyses. See Whitman et al. (2010) for a
recent meta-analysis on people’s wellbeing and firm
performance.
4 For the purpose of this chapter, we adopt a broad definition
of wellbeing, colloquially referred to as happiness, covering
evaluative measures such as overall life evaluation and
domain-specific job satisfaction, experiential measures
(both positive and negative affect), and eudemonic
measures (employee engagement).
5 The present discussion on the overall importance of
employment, as well as the state of job satisfaction and
employee engagement worldwide, in this chapter builds
to some extent on De Neve and Ward (2017).
6 The lower average life evaluation for the self-employed may
come as a surprise, but is in line with an emerging strand of
literature on the misprediction of wellbeing consequences
when deciding to become self-employed (Odermatt et al.,
2017). A possible mechanism may be that individuals who
become self-employed underestimate the associated rise in
workload. Moreover, as discussed in De Neve and Ward
(2017), the relationship between life evaluation and being
self-employed varies by world region.
7 A potential mechanism behind this finding is that the
previously unemployed are scared of becoming unemployed
again (Knabe and Raetzel, 2011).
8 For summaries of the work on the importance of being
in employment (and of being out of unemployment),
including differences by gender, for wellbeing, see also
What Works Centre for Wellbeing (2017a, 2017b).
9 See Table 1 in the Web Appendix.
10 Studying job satisfaction has a history in business
economics (see Spector (1997) or Cooper and Robertson
(2003), for example). While being more domain-specific
than overall life evaluation, this indicator is also more prone
to framing effects, as the relationship between wellbeing
and work is revealed to the respondent.
11 See Table W11 in the Web Appendix for summary statistics
of job satisfaction, the different elements of workplace
quality, and the demographic control variables, including
their definitions.
12 Tables W9 and W10 in the Web Appendix show differences
in average job satisfaction and workplace quality by region
in numbers; Table W9 shows these values for the different
domains, Table W10 for the different constituent elements
within each domain. Table W11 provides definitions and
summary statistics of the variable used. Table W12 gives
an overview of the different countries covered within
each region.
13 For a comprehensive summary of a systematic review on
the relationship between job quality and wellbeing, see
also What Works Centre for Wellbeing (2017c).
14 The important role of purpose for performance has also
been studied in educational contexts: Yeager et al. (2014)
show that promoting a pro-social, self-transcendent
purpose improves academic self-regulation in students.
15 The company later offered the option to work from home
to the whole firm, allowing formerly treated employees to
re-select between working from home or working in the
office: about half of them switched back, which almost
doubled performance gains to 22%. This highlights the
importance of accounting for self-selection and learning.
In fact, in a recent discrete choice experiment, Mas and
Pallais (2017) demonstrate that employee preferences for
flexible work practices are quite heterogeneous: while
most employees prefer a little extra income over flexibility,
to a small number of employees, flexible work practices
are very important.
16 On the importance of learning on the job for wellbeing,
see also What Works Centre for Wellbeing (2017d).
17 On the importance of team work more generally for
wellbeing, see What Works Centre for Wellbeing (2017e).
18 Note that pro-social behavior is distinct from altruism in that
it is not purely motivated by increasing another individual’s
welfare, but can be motivated by, for example, empathy,
reciprocity, or self-image (Evren and Minardi, 2017).
19 Following recommendations by Dolan and Metcalfe (2012),
the Office for National Statistics (ONS) in the UK now
routinely asks people how they think and feel about their
lives, including four items on evaluative (life satisfaction),
experiential (happiness, anxiousness), and eudemonic
(worthwhileness) measures of subjective wellbeing in
its surveys.

References
Amabile, T. M., Creativity in Context, Boulder, Westview, 1996.
Amabile, T. M., R. Conti, H. Coon, J. Lazenby, and M. Herron,
“Assessing the Work Environment for Creativity,” Academy of
Management Journal, 39(5), 1154-1184, 1996.
Amabile, T. M., and S. Kramer, The Progress Principle: Using
Small Wins to Ignite Joy, Engagement, and Creativity at Work,
Cambridge, Harvard University Press, 2011.
Angrave, D., and A. Charlwood, “What is the relationship
between long working hours, over-employment, underemploy-
ment and the subjective well-being of workers? Longitudinal
evidence from the UK,” Human Relations, 68(9), 1491-1515, 2015.
Anik, L., L. B. Aknin, M. I. Norton, E. W. Dunn, and J. Quoidbach,
“Prosocial Bonuses Increase Employee Satisfaction and Team
Performance,” PLoS ONE, 8(9), e75509, 2013.
Arends, I., C. Prinz, and F. Abma, “Job quality, health and
at-work productivity,” OECD Social, Employment and Migration
Working Papers, 195, 2017.
Ariely, D., E. Kamenica, and D. Prelec, “Man’s search for
meaning: The case of Legos,” Journal of Economic Behavior &
Organization, 67, 671-677, 2008.
Artz, B. M., A. H. Goodall, and A. J. Oswald, “Boss Competence
and Worker Well-Being,” Industrial and Labor Relations Review,
70(2), 419-450, 2017.
Ashraf, N., and O. Bandiera, “Altruistic Capital,” American
Economic Review: Papers & Proceedings, 107(5), 70-75, 2017.
Batson, C. D., and A. A. Powell, “Altruism and Prosocial
Behavior,” in Weiner, I. B. (Ed.), Handbook of Psychology
Volume 5, London, Wiley, 2003.
Bloom, N., J. Liang, J. Roberts, and Z. J. Ying, “Does Working
from Home Work? Evidence from a Chinese Experiment,”
Quarterly Journal of Economics, 130, 165-218, 2015.
Bockerman, P., and P. Ilmakunnas, “The Job Satisfaction-
Productivity Nexus: A Study Using Matched Survey and
Register Data,” Industrial and Labor Relations Review, 65(2),
244-262, 2012.
Boeheim, R., and M. P. Taylor, “Actual and Preferred Working
Hours,” British Journal of Industrial Relations, 42(1), 149-166,
2004.
Bohnet, I., A. van Green, and M. Bazerman, “When Performance
Trumps Gender Bias: Joint vs. Separate Evaluation,” Management
Science, 62(5), 1225-1234, 2015.
Bryson, A., and G. MacKerron, “Are You Happy While You
Work?,” Economic Journal, 127(599), 106-125, 2017.
Chandola, T., E. Brunner, and M. Marmot, “Chronic stress at
work and the metabolic syndrome: prospective study,” British
Medical Journal, 332(7540), 521-525, 2006.
Clark, A., “Work, Jobs and Well-Being across the Millennium,”
IZA Discussion Paper, 3940, 2009.
Clark, A. E., “Unemployment as a Social Norm: Psychological
Evidence from Panel Data,” Journal of Labor Economics, 21(2),
323-351, 2003.
Clark, A. E., and A. J. Oswald, “Unhappiness and Unemployment,”
Economic Journal, 104(424), 648-659, 1994.
Clark, A. E., and Y. Georgellis, “Back to Baseline in Britain:
Adaptation in the British Household Panel Survey,” Economica,
80(319), 496-512, 2013.
Clark, A. E., E. Diener, Y. Georgellis, and R. Lucas, “Lags and
Leads in Life Satisfaction: A Test of the Baseline Hypothesis,”
Economic Journal, 118(529), F222-F243, 2008.
Clark, A. E., Y. Georgellis, and P. Sanfey, “Scarring: The
Psychological Impact of Past Unemployment,” Economica,
68(270), 221-241, 2001.
Cooper, C., and I. T. Robertson, Management and Happiness,
Cheltenham, Edward Elgar, 2013.
Danziger, S., J. Levav, and L. Avnaim-Pesso, “Extraneous factors
in judicial decisions,” Proceedings of the National Academy of
Sciences, 108(17), 6889-6892, 2011.
De Neve, J.-E., and A. J. Oswald, “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, 2012.
De Neve, J.-E., E. Diener, L. Tay, and C. Xuereb, “The Objective
Benefits of Subjective Well-Being,” in World Happiness Report
2013, edited by J. Helliwell, R. Layard, and J. Sachs.
De Neve, J.-E., and G. Ward, “Happiness at Work,” in World
Happiness Report 2017, edited by J. Helliwell, R. Layard, and J.
Sachs.
Dolan, P. H., “Happiness by Design: Change What You Do, Not
How You Think,” New York: Hudson Street Press.
Dolan, P. H., and R. Metcalfe, “Measuring Subjective Wellbeing:
Recommendations on Measures for Use by National
Governments,” Journal of Social Policy, 41(2), 409-427.
Dunn, E. W., L. B. Aknin, and M. I. Norton, “Spending Money on
Others Promotes Happiness,” Science, 319(5870), 1687-1688,
2008.
Dur, R., C. Bradler, S. Neckermann, and A. Non, “Employee
Recognition and Performance: A Field Experiment,” Management
Science, 62(11), 3085-3099, 2016.
Edmans, A., “Does the Stock Market Fully Value Intangibles?
Employee Satisfaction and Equity Prices,” Journal of Financial
Economics, 101(3), 621-640, 2011.
Edmans, A., “The Link Between Job Satisfaction and Firm
Value, With Implications for Corporate Social Responsibility,”
Academy of Management Perspectives, 26(4), 1-19, 2012.
Eisenberg, N., T. L. Spinrad, and A. S. Morris, “Prosocial
Development,” in: Zelazo, P. D. (Ed.), Oxford Handbook of
Developmental Psychology Volume 2, Oxford, Oxford University
Press, 2013.
Evren, O., and S. Minardi, “Warm-glow Giving and Freedom to
be Selfish,” Economic Journal, 127(603), 1381-1409, 2017.
Fehr, E., and U. Fischbacher, “The nature of human altruism,”
Nature, 425(6960), 785-791, 2003.
Frey, B. S., “Research on Well-Being: Determinants, Effects, and its
Relevance for Management,” CREMA Working Paper, 2017-11, 2017.
Frey, B. S., and J. Gallus, Honour versus Money. The Economics
of Awards, Oxford, Oxford University Press, 2017.
Gallup News, “Employees Want a Lot More From Their
Managers,” Online: http://news.gallup.com/businessjournal/
182321/employees-lot-managers.aspx, accessed 01/12/2017,
2015.
Gallus, J., and B. S. Frey, “Awards: A Strategic Management
Perspective,” Strategic Management Journal, 37(8), 1699-1714,
2016.
Global Happiness Policy Report 2018

110
111
Gartenberg, C., A. Prat, and G. Serafeim, “Corporate Purpose
and Financial Performance,” Harvard Business School Working
Paper, 17-023, 2016.
Gielen, A. C., and J. C. van Ours, “Unhappiness and Job
Finding,” Economica, 81(323), 544-565, 2014.
Gilchrist, D. S., M. Luca, and D. Malhotra, “When 3 + 1 > 4: Gift
Structure and Reciprocity in the Field,” Management Science,
62(9), 2639-2650, 2016.
Graham, C., “Happiness and Economics: Insights for Policy from
a new Science of Well-being,” Journal of Behavioral Economics
for Policy, 1(1), 69-72, 2017.
Grant, A. M., “Employees without a Cause: The Motivational
Effects of Prosocial Impact in Public Service,” International
Public Management Journal, 11(1), 48-66, 2008.
Harrison, D. A., D. A. Newman, and P. L. Roth, “How Important
are Job Attitudes? Meta-Analytic Comparisons of Integrative
Behavioral Outcomes and Time Sequences,” Academy of
Management Journal, 49(2), 305-325, 2006.
Harter, J. K., F. L. Schmidt, J. W. Asplund, E. A. Killham, and S.
Agrawal, “Causal Impact of Employee Work Perceptions on the
Bottom Line of Organizations,” Perspectives on Psychological
Science, 5(4), 378-389, 2010.
Harter, J. K., F. L. Schmidt, T. L. Hayes, “Business-Unit-Level
Relationship Between Employee Satisfaction, Employee
Engagement, and Business Outcomes: A Meta-Analysis,”
Journal of Applied Psychology, 87(2), 268-279, 2002.
Heinz, M., S. Jeworrek, V. Mertins, H. Schumacher, and M. Sutter,
“Measuring Indirect Effects of Unfair Employer Behavior on
Worker Productivity—A Field Experiment,” Max Planck Institute
for Research on Collective Goods Working Paper, 2017/22, 2017.
Helliwell, J. F., and H. Huang, “Well-Being and Trust in the
Workplace,” Journal of Happiness Studies, 12(5), 747-767, 2011.
Helliwell, J. F., and S. Wang, “How Was the Weekend? How the
Social Context Underlies Weekend Effects in Happiness and Other
Emotions for US Workers,” PLoS ONE, 10(12), e014512, 2015.
Helliwell, J. F., H. Huang, and R. D. Putnam, “How’s the Job? Are
Trust and Social Capital Neglected Workplace Investments?,” in
Bartkus, V. O., and J. H. Davis (Ed.), Social Capital: Reaching
Out, Reaching In, London, Edward Elgar, 2009.
Helliwell, J. F., H. Huang, and S. Wang, “New Evidence on Trust
and Wellbeing,” NBER Working Paper, 22450, 2016.
Helliwell, J. F., L. B. Aknin, H. Shiplett, H. Huang, and S. Wang,
“Social Capital and Prosocial Behavior as Sources of Well-
Being,” NBER Working Paper, 23761, 2017.
Hu, J., and J. B. Hirsh, “Accepting Lower Salaries for Meaningful
Work,” Frontiers in Psychology, 8(1649), 1-10, 2017.
Judge, T. A., C. J. Thoresen, J. E. Bono, and G. K. Patton, “The
job satisfaction-job performance relationship: A qualitative and
quantitative review,” Psychological Bulletin, 127(3), 376-407,
2001.
Kahneman, D., A. B. Krueger, D. A. Schkade, N. Schwarz, and A.
A. Stone, “A Survey Method for Characterizing Daily Life
Experience: The Day Reconstruction Method,” Science,
306(5702), 1776-1780, 2004.
Kassenboehmer, S. C., and J. P. Haisken-DeNew, “You’re Fired!
The Causal Negative Effect of Entry Unemployment on Life
Satisfaction,” Economic Journal, 119(536), 448-462, 2008.
Kelly, E. L., P. Moen, J. M. Oakes, W. Fan, C. Okechukwu, K. D.
Davis, L. B. Hammer, E. E. Kossek, R. B. King, G. C. Hanson, F.
Mierzwa, and L. M. Casper, “Changing Work and Work-Family
Conflict: Evidence from the Work, Family, and Health Network,”
American Sociological Review, 79(3), 485-516, 2014.
King, R. B., G. Karuntzos, L. M. Casper, P. E. Moen, K. D. Davis, L.
F. Berkman, M. Durham, and E. E. Kossek, “Work-Family Balance
Issues and Work-Leave Policies,” in Gatchel, R. J., and I. Z.
Schultz (Eds.), Handbook of Occupational Health and Wellness,
New York, Springer, 2012.
Knabe, A., and S. Raetzel, “Scarring or Scaring? The Psychological
Impact of Past Unemployment and Future Unemployment
Risk,” Economica, 78(310), 283-293, 2011.
Knabe, A., S. Raetzel, R. Schoeb, and J. Weimann, “Dissatisfied
with Life but Having a Good Day: Time-use and Well-being of
the Unemployed,” Economic Journal, 120(547), 867-889, 2010.
Knight, C., and S. A. Haslam, “The Relative Merits of Lean,
Enriched, and Empowered Offices: An Experimental Examination
of the Impact of Workspace Management Strategies on
Well-Being and Productivity,” Journal of Experimental Psychology:
Applied, 16(2), 158-172, 2010.
Krause, A., “Don’t worry, be happy? Happiness and reemployment,”
Journal of Economic Behavior & Organization, 96, 1-20, 2013.
Krueger, A. B., and A. I. Mueller, “Time Use, Emotional Well-
Being, and Unemployment: Evidence from Longitudinal Data,”
American Economic Review: Papers & Proceedings, 102(3),
594-599, 2012.
Kunze, L., and N. Suppa, “Bowling alone or bowling at all? The
effect of unemployment on social participation,” Journal of
Economic Behavior & Organization, 133, 213-235, 2017.
Lazear, E. P., K. L. Shaw, and C. T. Stanton, “The Value of
Bosses,” Journal of Labor Economics, 33(4), 823-861, 2015.
Leslie, L. M., C. Flaherty Manchester, T.Y. Park, and S. Ahn Meng,
“Flexible Work Practices: A Source of Career Premiums or
Penalties?,” Academy of Management Journal, 55(6), 1407-1428,
2012.
Luechinger, S., S. Meier, and A. Stutzer, “Why Does Unemployment
Hurt the Employed? Evidence from the Life Satisfaction Gap
Between the Public and the Private Sector,” Journal of Human
Resources, 45(4), 998-1045, 2010.
Marcus, J., “The effect of unemployment on the mental health
of spouses—Evidence from plant closures in Germany,” Journal
of Health Economics, 32(3), 546-558, 2013.
Mas, A., and A. Pallais, “Valuing Alternative Work Arrange-
ments,” American Economic Review, 107(12), 3722-3759, 2017.
Meier, S., and A. Stutzer, “Is Volunteering Rewarding in Itself?,”
Economica, 75(297), 39-59, 2008.
Moen, P., E. L. Kelly, and R. Hill, “Does Enhancing Work-Time
Control and Flexibility Reduce Turnover? A Naturally Occurring
Experiment,” Social Problems, 58(1), 69-98, 2011.
Moen, P., E. L. Kelly, W. Fan, S.R. Lee, D. Almeida, E. E. Kossek,
and O. Buxto, “Does a Flexibility/Support Organizational
Initiative Improve High-Tech Employees’ Well-Being? Evidence
from the Work, Family, and Health Network,” American
Sociological Review, 81(1), 134-164, 2016.
OECD, “Balancing paid work, unpaid work and leisure,” Online:
http://www.oecd.org/gender/data/balancingpaidworkunpaid-
workandleisure.htm, accessed 08/10/2017, 2014.

Global Happiness Policy Report 2018
OECD, “Employment: Time spent in paid and unpaid work, by
sex,” Online: http://stats.oecd.org/index.aspx?queryid=54757,
accessed 19/10/2017, 2017a.
OECD, “OECD Guidelines on Measuring the Quality of
the Working Environment,” Online: http://www1.oecd.org/
publications/oecd-guidelines-on-measuring-the-quality-of-the-
working-environment-9789264278240-en.htm, accessed
24/11/2017, 2017b.
Ogbonnaya, C., and K. Daniels, “Good Work, Wellbeing and
Changes in Performance Outcomes: Illustrating the Effects of
Good People Management Practices with an Analysis of the
National Health Service,” mimeo, 2017.
Oswald, A. J., E. Proto, and D. Sgroi, “Happiness and Productivity,”
Journal of Labor Economics, 33(4), 789-822, 2015.
PwC Research, “The Rising Cost of Absence,” 2013.
Sainsbury Centre for Mental Health, “Mental health at work:
developing the business case,” 2007.
Social Research and Demonstration Corporation, “UPSKILL: A
Credible Test of Workplace Literacy and Essential Skills
Training—Summary Report,” Online: http://www.srdc.org/
media/ 199770/upskill-final-results-es-en.pdf, accessed
29/11/2017, 2014a.
Social Research and Demonstration Corporation, “UPSKILL: A
Credible Test of Workplace Literacy and Essential Skills
Training—Technical Report,” Online: http://www.srdc.org/
media/199774/upskill-technical-report-en.pdf, accessed
29/11/2017, 2014b.
Spector, P. E., Job Satisfaction: Application, Assessment,
Causes, and Consequences, Thousand Oaks, Sage, 1997.
Stevenson, B., and J. Wolfers, “The Paradox of Declining Female
Happiness,” American Economic Journal: Economic Policy, 1(2),
190-225, 2009.
Tay, L., and J. K. Harter, “Economic and Labor Market Forces
Matter for Worker Well-Being,” Applied Psychology: Health and
Well-Being, 5(2), 193-208, 2013.
Tenney, E. R., J. M. Poole, and E. Diener, “Does Positivity
Enhance Work Performance? Why, When, and What We Don’t
Know,” Research on Organizational Behaviors, 36, 27-46, 2016.
Tims, M., A. B. Bakker, and D. Derks, “The Impact of Job Crafting
on Job Demands, Job Resources, and Well-Being,” Journal of
Occupational Health Psychology, 18(2), 230-240, 2013.
van Praag, B. M. S., P. Frijters, and A. Ferreri-Carbonell, “The
anatomy of subjective well-being,” Journal of Economic
Behavior & Organization, 51(1), 29-49, 2003.
What Works Centre for Wellbeing, “Briefing: Unemployment,
(Re)Employment and Wellbeing,” Online: https://www.
whatworkswellbeing.org/product/unemployment-reemploy-
ment-and-wellbeing/, accessed 27/11/2017, 2017a.
What Works Centre for Wellbeing, “Briefing: Gender and
Wellbeing,” Online: https://www.whatworkswellbeing.org/
product/gender-and-wellbeing/, accessed 27/11/2017, 2017b.
What Works Centre for Wellbeing, “Briefing: Job Quality and
Wellbeing,” Online: https://www.whatworkswellbeing.org/
product/job-quality-and-wellbeing/, accessed 27/11/2017, 2017c.
What Works Centre for Wellbeing, “Briefing: Learning at Work
and Wellbeing,” Online: https://www.whatworkswellbeing.org/
product/learning-at-work/, accessed 27/11/2017, 2017d.
What Works Centre for Wellbeing, “Briefing: Team Working,”
Online: https://www.whatworkswellbeing.org/product/
team-working/, accessed 27/11/2017, 2017e.
White, M., and A. Bryson, “Positive Employee Relations: How
Much Human Resource Management Do You Need?,” Human
Relations, 66(3), 385-406, 2013.
Whitman, D. S., D. L. Van Rooy, and C. Viswesvaran, “Satisfaction,
citizenship behaviors, and performance in work units: A
meta-analysis of collective construct relations,” Personnel
Psychology, 63(1), 41-81, 2010.
Willis Towers Watson, “Balancing Employer and Employee
Priorities: Insights From the 2014 Global Workforce and Global
Talent Management and Rewards Studies,” Online: https://
www.towerswatson.com/en/Insights/IC-Types/Survey-
Research-Results/2014/07/balancing-employer-and-employ-
ee-priorities, accessed 01/12/2017, 2014.
Winkelmann, L., and R. Winkelmann, “Why are the Unemployed
So Unhappy? Evidence from Panel Data,” Economica, 65(257),
1-15, 1998.
Wooden, M., D. Warren, and R. Drago, “Working Time Mismatch
and Subjective Well-being,” British Journal of Industrial
Relations, 47(1), 147-179, 2009.
Woolley, K., and A. Fishbach, “The Experience Matters More
Than You Think: Weighting Intrinsic Incentives More Inside Than
Outside of an Activity,” Journal of Personality and Social
Psychology, 109(6), 968-982, 2015.
Wrzesniewski, A., and J. E. Dutton, “Crafting a Job: Revisioning
Employees as Active Crafters of Their Work,” Academy of
Management Review, 26(2), 179-201, 2001.
Wunder, C., and G. Heineck, “Working time preferences, hours
mismatch and well-being of couples: Are there spillovers?,”
Labor Economics, 24, 244-252, 2013.
Yeager, D. S., M. D. Henderson, S. D’Mello, D. Paunesku, G. M.
Walton, B. J. Spitzer, and A. L. Duckworth, “Boring but
Important: A Self-Transcendent Purpose for Learning Fosters
Academic Self-Regulation,” Journal of Personality and Social
Psychology, 107(4), 559-580, 2014.
Yuan, L., “The Happier One Is, the More Creative One Becomes:
An Investigation on Inspirational Positive Emotions from Both
Subjective Well-Being and Satisfaction at Work,” Psychology, 6,
201-209, 2015.

112
113
Appendix

Global Happiness Policy Report 2018
Table W1: Importance of Domain Satisfactions for Life Satisfaction
Life
Satisfaction
Level Effects Workers Non-Workers
Domain Satisfaction Rank of
Importance
All West East All West East
Financial Satisfaction 10,7480 0,6370 0,8590 0,8280 0,7600 0,8960
Health Satisfaction 20,4730 0,5010 0,4450 0,6585 0,6580 0,6590
Job Satisfaction 30,3905 0,3520 0,4290 - - -
Leisure Satisfaction 40,2465 0,2240 0,2690 0,3585 0,1820 0,5350
House Satisfaction 50,1660 0,1480 0,1840 0,2635 0,2000 0,3270
Environment Satisfac-
tion
60,1370 0,0500 0,2240 0,1885 0,0660 0,3110
Notes: Level effects obtained from ordered probit models with individual random effects, adapted from van Praag et
al. (2003). The authors regress life satisfaction on different domain satisfactions of respondents, controlling for year
dummies, mean domain satisfactions, age, gender, partnership status, years of education, household income,
available leisure time, mean household income, and mean available leisure time. The respective level effect is
calculated as the sum of the individual domain satisfaction and the mean of that domain satisfaction.
Source: German Socio-Economic Panel, Years 1992 to 1997

114
115
Table W2: Subjective Well-being and Employment Status
(Gallup World Poll, 2014-2016)
Cantril Ladder Positive Affect Negative Affect
Employment (v. employed FT for employer)
Employed FT for Self -0.019** -0.008 0.019**
(0.008) (0.009) (0.009)
Employed PT (does not want FT) 0.058*** 0.010 -0.013
(0.010) (0.010) (0.010)
Employed PT (wants FT) -0.087*** -0.006 0.093***
(0.009) (0.009) (0.011)
Unemployed -0.221*** -0.120*** 0.221***
(0.011) (0.012) (0.013)
Out of Workforce -0.037*** -0.062*** 0.021**
(0.008) (0.008) (0.010)
Covariates
Income 0.211*** 0.121*** -0.132***
(0.006) (0.005) (0.005)
Eduction: Medium 0.158*** 0.089*** -0.095***
(0.008) (0.010) (0.011)
Eduction: High 0.310*** 0.199*** -0.136***
(0.011) (0.013) (0.014)
Marital Status: Married/Domestic Partner 0.051*** 0.007 -0.024***
(0.007) (0.006) (0.007)
Marital Status: Divorced/Separated -0.089*** -0.119*** 0.131***
(0.010) (0.010) (0.010)
Marital Status: Widowed -0.104*** -0.133*** 0.158***
(0.014) (0.014) (0.014)
Female 0.084*** 0.015*** 0.079***
(0.006) (0.005) (0.007)
Age -0.020*** -0.025*** 0.021***
(0.002) (0.002) (0.002)
Age^2 0.000*** 0.000*** -0.000***
(0.000) (0.000) (0.000)
Children in Household -0.022*** -0.011** 0.039***
(0.005) (0.006) (0.005)
Adults in Household -0.009*** -0.005*** 0.011***
(0.002) (0.002) (0.002)
Country + Year FEs Yes Ye s Yes
Observations 309263 288041 288041
R-squared 0.076 0.031 0.035
Countries 154 153 153
Country-Years 417 417 417
Standard errors clustered at country level in parantheses
* p<0.1 ** p<0.05

Global Happiness Policy Report 2018
Table W3: Effect of Workplace Quality on Job Satisfaction,
Aggregated Domains (Regression Equivalent to Figure 8)
Workplace Quality Effect on Job Satisfaction
Pay 0.131***
(0.0161)
Working Hours -0.0107
(0.0104)
Working Hours Mismatch -0.0271**
(0.0106)
Work-Life Imbalance -0.106***
(0.00681)
Skills Match 0.0474***
(0.00880)
Job Security 0.0734***
(0.00906)
Difficulty, Stress, Danger -0.0918***
(0.0105)
Opportunities for Advancement 0.0598***
(0.0119)
Independence 0.0551***
(0.0109)
Interesting Job 0.267***
(0.0231)
Interpersonal Relationships 0.281***
(0.0145)
Usefulness 0.0399***
(0.0103)
Union Member -0.00322
(0.00569)
Age -0.116**
(0.0435)
Age Squared 0.147***
(0.0418)
Female 0.00505
(0.00731)
Partnered 0.0357***
(0.0100)
Separated 0.0145**
(0.00591)
Divorced 0.0134**
(0.00606)
Widowed 0.00639
(0.00701)
Years of Education -0.0569***
(0.00831)
Number of Individuals in Household -0.0126
(0.0111)
Number of Children in Household 0.00535
(0.0129)
Number of Toddlers in Household 0.00302
(0.00940)
Constant -0.0439
(0.206)
Occupation Fixed Effects Ye s
Industry Fixed Effects Yes
Country Fixed Effects Yes
Observations 16,326
Adjusted R-Squared 0.422
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation one; regressors are thus beta
coefficients. Squaring a regressor yields the respective share in the variation of job satisfaction that this regressor explains. Pay, Working Hours
Mismatch, Work-Life Imbalance, Skills Match, Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle
components obtained from separate principle component analyses that condense various variables in the respective domain of workplace quality into
a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web Appendix for summary statistics of the variables. The
sample is restricted to all individuals who state that they are working and who report working hours greater than zero.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

116
117
Table W4: Effect of Workplace Quality on Job Satisfaction, Disaggregated Domains
Workplace Quality Effect on
Job Satisfaction
High Income 0.0866***
(0.0122)
Individual Income (Natural Log) 0.105**
(0.0506)
Working Hours (Natural Log) -0.0105
(0.00980)
Wants to Work Same Hours Reference Category
Wants to Work More Hours -0.00979
(0.00697)
Wants to Work Less Hours -0.0297***
(0.00996)
Working on Weekends 0.0169**
(0.00699)
Work Interfering With Family -0.109***
(0.00935)
Difficulty of Taking Time Off -0.0385***
(0.00900)
Skills Match 0.0476***
(0.00920)
Skills Training 0.0190**
(0.00878)
Job Security 0.0700***
(0.00847)
Hard Physical Work -0.00739
(0.0119)
Stressful Work -0.0853***
(0.0113)
Opportunities for Advancement 0.0538***
(0.0114)
Independent Work 0.0275**
(0.0106)
Working From Home -0.00996
Daily Work Flexible Reference Category
Daily Work Fixed -0.0112
(0.00846)
Daily Work Free 0.0386***
(0.0100)
Working Hours Flexible Reference Category
Working Hours Fixed -0.00195
(0.00742)
Working Hours Free -0.00270
(0.00835)
Working Schedule Flexible Reference Category
Working Schedule Fixed -0.0212**
(0.00798)
Working Schedule Free -0.0167**
(0.00793)
Interesting Job 0.265***
(0.0221)
Contact With Other People 0.00561
(0.00891)
Relationship With Management 0.222***
(0.0114)
Relationship With Co-Workers 0.0906***
(0.0116)
Helping Other People 0.0256***
(0.00901)
Being Useful to Society 0.0359***
(0.00853)
Union Member -3.23e-05
(0.00635)
Age -0.0938**
(0.0430)
Age Squared 0.121***
(0.0401)
Female 0.0102
(0.00752)
Partnered 0.0375***
(0.00953)
Separated 0.0171***
(0.00600)
Divorced 0.0135**
(0.00604)
Widowed 0.00955
(0.00693)
Years of Education -0.0445***
(0.00880)
Number of Individuals in
Household
-0.0120
(0.0110)
Number of Children in Household 0.00896
(0.0136)
Number of Toddlers in Household 0.00560
(0.00953)
Constant 0.0461
(0.209)
Occupation Fixed Effects Ye s
Industry Fixed Effects Ye s
Country Fixed Effects Yes
Observations 16,326
Adjusted R-Squared 0.438
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation one; regressors are thus
beta coefficients. Squaring a regressor yields the respective share in the variation of job satisfaction that this regressor explains. The sample is
restricted to all individuals who state that they are working and who report working hours greater than zero.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

Global Happiness Policy Report 2018
Table W5: Effect of Workplace Quality on Job Satisfaction, by Employment
Status (Regression Equivalent to Figure 9a)
Effect on Job Satisfaction
Workplace Quality Employed Self-Employed
Pay 0.134*** 0.153***
(0.0157) (0.0401)
Working Hours -0.00159 -0.0156
(0.0104) (0.0164)
Working Hours Mismatch -0.0343*** 0.0221
(0.00760) (0.0331)
Work-Life Imbalance -0.109*** -0.0666***
(0.00786) (0.0237)
Skills Match 0.0474*** 0.0555**
(0.0101) (0.0249)
Job Security 0.0672*** 0.104***
(0.0101) (0.0257)
Difficulty, Stress, Danger -0.102*** -0.0458*
(0.0119) (0.0236)
Opportunities for
Advancement
0.0611*** 0.0297
(0.0122) (0.0310)
Independence 0.0469*** 0.0899**
(0.00983) (0.0333)
Interesting Job 0.264*** 0.246***
(0.0236) (0.0401)
Interpersonal Relation-
ships
0.295*** 0.156***
(0.0135) (0.0376)
Usefulness 0.0407*** 0.0270
(0.0111) (0.0223)
Union Member 0.00138 -0.0197
(0.00656) (0.0241)
Age -0.137*** 0.130
(0.0425) (0.171)
Age Squared 0.165*** -0.0920
(0.0427) (0.150)
Female 0.00534 0.0276
(0.00721) (0.0273)
Partnered 0.0361*** 0.0364
(0.00972) (0.0362)
Separated 0.0195*** -0.0203
(0.00664) (0.0188)
Divorced 0.0160** 0.0114
(0.00740) (0.0279)
Widowed 0.00709 0.0209
(0.00716) (0.0226)
Years of Education -0.0592*** -0.02 02
(0.0105) (0.0237)
Number of Individuals
in Household
-0.0109 -0.00464
(0.0119) (0.0246)
Number of Children in
Household
0.00596 -0.0150
(0.0114) (0.0318)
Number of Toddlers
in Household
0.00118 0.00632
(0.00824) (0.0296)
Constant -0.0223 -1.228***
(0.212) (0.216)
Occupation Fixed Effects Yes Ye s
Industry Fixed Effects Yes Ye s
Country Fixed Effects Ye s Yes
Observations 14,113 2,059
Adjusted R-Squared 0.437 0.291
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation one; regressors are thus beta
coefficients. Squaring a regressor yields the respective share in the variation of job satisfaction that this regressor explains. Pay, Working Hours
Mismatch, Work-Life Imbalance, Skills Match, Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle
components obtained from separate principle component analyses that condense various variables in the respective domain of workplace quality into
a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web Appendix for summary statistics of the variables. The
sample is restricted to all individuals who state that they are working and who report working hours greater than zero.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

118
119
Table W6: Effect of Workplace Quality on Job Satisfaction, by Working Time
(Regression Equivalent to Figure 9b)
Effect on Job Satisfaction
Workplace Quality Full-Time Part-Time
Pay 0.128*** 0.164***
(0.0195) (0.0308)
Working Hours 0.0316 0.00650
(0.0216) (0.0118)
Working Hours Mismatch -0.0301*** -0.000893
(0.00977) (0.0281)
Work-Life Imbalance -0.120*** -0.0682***
(0.00973) (0.0200)
Skills Match 0.0428*** 0.0773***
(0.00924) (0.0200)
Job Security 0.0730*** 0.0714***
(0.00976) (0.0174)
Difficulty, Stress, Danger -0.0883*** -0.101***
(0.0110) (0.0214)
Opportunities for
Advancement
0.0628*** 0.0284
(0.0125) (0.0217)
Independence 0.0526*** 0.0588***
(0.0117) (0.0201)
Interesting Job 0.255*** 0.312***
(0.0247) (0.0247)
Interpersonal Relation-
ships
0.291*** 0.234***
(0.0148) (0.0259)
Usefulness 0.0407*** 0.0444*
(0.0112) (0.0239)
Union Member -0.00166 -0.00716
(0.00617) (0.0211)
Age -0.0488 -0.309**
(0.0469) (0.125)
Age Squared 0.0832 0.311**
(0.0500) (0.119)
Female 0.00946 -0.0123
(0.00786) (0.0150)
Partnered 0.0272*** 0.0757**
(0.00970) (0.0281)
Separated 0.0111 0.0229*
(0.00666) (0.0128)
Divorced 0.00996 0.0322
(0.00662) (0.0229)
Widowed 0.00630 0.0230
(0.00727) (0.0206)
Years of Education -0.0518*** -0.0610**
(0.0102) (0.0276)
Number of Individuals in
Household
-0.00608 -0.00405
(0.0115) (0.0271)
Number of Children in
Household
0.00269 -0.00480
(0.0146) (0.0202)
Number of Toddlers in
Household
-0.00460 0.00468
(0.00916) (0.0276)
Constant -0.239 0.0832
(0.214) (0.496)
Occupation Fixed Effects Yes Ye s
Industry Fixed Effects Yes Ye s
Country Fixed Effects Ye s Yes
Observations 13,345 2,981
Adjusted R-Squared 0.430 0.392
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation one; regressors are thus beta
coefficients. Squaring a regressor yields the respective share in the variation of job satisfaction that this regressor explains. Pay, Working Hours
Mismatch, Work-Life Imbalance, Skills Match, Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle
components obtained from separate principle component analyses that condense various variables in the respective domain of workplace quality into
a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web Appendix for summary statistics of the variables. The
sample is restricted to all individuals who state that they are working and who report working hours greater than zero. Full-Time: work-ing at least 35
hours per week, Part-Time: working less than 35 hours per week.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

Global Happiness Policy Report 2018
Table W7: Effect of Workplace Quality on Job Satisfaction, by Gender
(Regression Equivalent to Figure 9c)
Effect on Job Satisfaction
Workplace Quality Male Female
Pay 0.119*** 0.148***
(0.0254) (0.0157)
Working Hours - 0.0176 -0.00334
(0.0178) (0.00905)
Working Hours Mismatch -0.0149 -0.0365***
(0.0151) (0.00914)
Work-Life Imbalance -0.109*** -0.101***
(0.0127) (0.00941)
Skills Match 0.0478*** 0.0462***
(0.00937) (0.0144)
Job Security 0.0794*** 0.0691***
(0.00815) (0.0139)
Difficulty, Stress, Danger -0.0795*** -0.107***
(0.0150) (0.0134)
Opportunities for
Advancement
0.0629*** 0.0564***
(0.0148) (0.0156)
Independence 0.0626*** 0.0404***
(0.0149) (0.0119)
Interesting Job 0.256*** 0.276***
(0.0257) (0.0241)
Interpersonal
Relationships
0.286*** 0.278***
(0.0192) (0.0154)
Usefulness 0.0347*** 0.0431***
(0.0122) (0.0156)
Union Member -0.00579 0.00230
(0.00712) (0.0124)
Age -0.149*** -0.0907
(0.0537) (0.0735)
Age Squared 0.178*** 0.128*
(0.0525) (0.0694)
Partnered 0.0250** 0.0466***
(0.0118) (0.0128)
Separated 0.0182** 0.0118
(0.00806) (0.00869)
Divorced 0.0140 0.0141
(0.00965) (0.00930)
Widowed 0.00242 0.0101
(0.0114) (0.00968)
Years of Education -0.0564*** -0.0595***
(0.0113) (0.0113)
Number of Individuals
in Household
-0.0158 -0.00415
(0.0202) (0.0140)
Number of Children
in Household
0.00531 0.0108
(0.0174) (0.0145)
Number of Toddlers
in Household
0.000127 0.00819
(0.0134) (0.0118)
Constant 0.0498 -0.586
(0.215) (0.583)
Occupation Fixed Effects Yes Ye s
Industry Fixed Effects Yes Ye s
Country Fixed Effects Ye s Yes
Observations 8,405 7,921
Adjusted R-Squared 0.415 0.426
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation one; regressors are thus beta
coefficients. Squaring a regressor yields the respective share in the variation of job satisfaction that this regressor explains. Pay, Working Hours
Mismatch, Work-Life Imbalance, Skills Match, Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle
components obtained from separate principle component analyses that condense various variables in the respective domain of workplace quality into
a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web Appendix for summary statistics of the variables. The
sample is restricted to all individuals who state that they are working and who report working hours greater than zero.
Source: International Social Survey Program, Module on Work Orientations, Year 2015

120
121
Table W8: Effect of Workplace Quality on Job Satisfaction, by Education Level
(Regression Equivalent to Figure 9d)
Effect on Job Satisfaction
Workplace Quality Low Education Medium Education High Education
Pay 0.0232 0.153*** 0.121***
(0.0493) (0.0196) (0.0211)
Working Hours 0.0127 -0.0173 -0.00963
(0.0343) (0.0149) (0.0135)
Working Hours Mismatch 0.0284 -0.0253** -0.0325*
(0.0457) (0.00928) (0.0190)
Work-Life Imbalance -0.190*** -0.104*** -0.0898***
(0.0240) (0.0101) (0.0132)
Skills Match 0.0806 0.0606*** 0.0181
(0.0561) (0.0106) (0.0111)
Job Security 0.107** 0.0776*** 0.0588***
(0.0429) (0.0126) (0.0126)
Difficulty, Stress, Danger -0.0461 -0.0955*** -0.106***
(0.0395) (0.0108) (0.0202)
Opportunities for Advancement 0.0361 0.0570*** 0.0731***
(0.0331) (0.0153) (0.0170 )
Independence 0.0660 0.0536*** 0.0486***
(0.0463) (0.0115) (0.0164)
Interesting Job 0.298*** 0.234*** 0.324***
(0.0588) (0.0257) (0.0316)
Interpersonal Relationships 0.254*** 0.294*** 0.261***
(0.0657) (0.0141) (0.0201)
Usefulness -0.0329 0.0319*** 0.0625***
(0.0545) (0.0110) (0.0144)
Union Member -0.0417 -0.0103 0.00143
(0.0850) (0.00975) (0.0110)
Age 0.0333 -0.138** -0.0716
(0.261) (0.0610) (0.0888)
Age Squared 0.0480 0.171** 0.0916
(0.249) (0.0638) (0.0938)
Female -0.0579 0.00752 0.0103
(0.0550) (0.0115) (0.0125)
Partnered 0.00673 0.0449*** 0.0230*
(0.0582) (0.0131) (0.0132)
Separated -0.0450 0.0147 0.0185*
(0.0432) (0.00902) (0.0100)
Divorced -0.0157 0.0146** 0.00621
(0.0482) (0.00665) (0.0118)
Widowed -0.00566 0.00587 0.0150
(0.0305) (0.00846) (0.0135)
Years of Education -0.0621 -0.0498*** -0.0348***
(0.0590) (0.0130) (0.0127)
Number of Individuals in Household -0.0320 -0.00587 -0.0230*
(0.0530) (0.0131) (0.0126)
Number of Children in Household 0.0309 0.00440 0.00273
(0.0434) (0.0127) (0.0180)
Number of Toddlers in Household 0.0 521 -0.00559 0.00728
(0.0579) (0.0117) (0.0150)
Constant 0.791** 0.0212 -0.0144
(0.350) (0.397) (0.195)

Global Happiness Policy Report 2018
Effect on Job Satisfaction
Workplace Quality Low Education Medium Education High Education
Occupation Fixed Effects Ye s Yes Ye s
Industry Fixed Effects Ye s Ye s Ye s
Country Fixed Effects Ye s Yes Ye s
Observations 941 9,537 5,821
Adjusted R-Squared 0.314 0.425 0.442
Robust standard errors clustered at country level in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Notes: All variables (both left-hand side and right-hand side) are standardized with mean zero and standard deviation one; regressors are thus beta
coefficients. Squaring a regressor yields the respective share in the variation of job satisfaction that this regressor explains. Pay, Working Hours
Mismatch, Work-Life Imbalance, Skills Match, Difficulty, Stress, Danger, Independence, Interpersonal Relationships, and Usefulness are principle
components obtained from separate principle component analyses that condense various variables in the respective domain of workplace quality into
a single indicator; see Section 4 for a description of the procedure and Table W11 in the Web Appendix for summary statistics of the variables. The
sample is restricted to all individuals who state that they are working and who report working hours greater than zero. Low Education: highest degree
lower than secondary degree, Medium Education: highest degree secondary degree or vocational training, High Education: highest degree at least
lower tertiary degree.
Source: International Social Survey Program, Module on Work Orientations, Year 2015
Table W9: Average Job Satisfaction and Average Workplace Quality, by Region
AU + NZ CIS EA EL AC MENA NA SA SEA SSA
Job Satisfaction -0,1188 -0,0528 -0,3898 0,0079 0,2699 0,197 8 0,1125 0,0502 0,2262 -0,0691
Workplace Quality
Pay 0,0469 0,5131 0,5827 -0,1864 0,0420 0,0570 0,3608 0,0925 0,0687 -0,2175
Working Hours -0,3226 0,1493 0,1481 -0,0145 0,0932 -0,0763 -0,0212 0,0978 -0,0223 0,1354
Working Hours
Mismatch
0,1262 -0,4410 0,1188 0,0279 -0,1648 0,0023 -0,1209 -0,3452 -0,1336 -0,5507
Work-Life
Imbalance
0,0768 0,0013 0,0568 -0,038 9 -0,0021 -0,1474 0,0664 0,3821 0,3276 0,1021
Skills Match 0,2783 -0,6393 -0,2268 0,1046 -0,2969 -0,1168 0,3462 -0,5009 -0,3772 -0,2001
Job Security -0,1380 -0,0051 -0,1868 0,0021 0,0740 0,1414 0,1963 0,1 974 0,1940 0,0146
Difficulty, Stress,
Danger
0,0263 -0,0962 0,1314 -0,01 39 -0,1928 -0,2949 0,4442 0,6059 0,3322 0,0658
Opportunities for
Advancement
-0,0355 0,2047 -0,3084 -0,0598 0,2372 0,2177 0,3072 0,4015 0,5904 0,4213
Independence 0,0622 -0,5509 -0,0920 -0,0479 0,0102 0,1687 0,1430 0,3851 1,0541 -0,2453
Interesting Job -0,0 079 -0,1335 -0,5035 0,0634 0,0741 0,0673 0,1664 -0,4613 0,1549 -0,1434
Interpersonal
Relationships
0,0731 -0,4100 -0,2245 0,0239 0,1098 0,2435 0,0416 -0,2775 -0,1109 0,1693
Usefulness 0,0458 -0,2235 -0,1656 -0,0361 0,1 727 0,1422 0,3396 -0,2425 0,2147 -0,0260
Notes: All variables are standardized with mean zero and standard deviation one; negative values (marked in shades of red) indicate negative
deviations from the average value of the variable across countries, positive values (marked in shades of green) positive deviations. Observations
are weighted using country weights. The sample is restricted to all individuals who state that they are working and who report working hours
greater than zero. AU + NZ: Australia + New Zealand, CIS: Commonwealth of Independent States, EA: East Asia, E: Europe, LAC: Latin America and
Caribbean, MENA: Middle East and North Africa, NA: North America, SA: South Asia, SEA: South-East Asia, SSA: Sub-Saharan Africa.
Source: International Social Survey Program, Module on Work Orientations, Year 2015
Table W8 continued

122
123
Table W10: Average Job Satisfaction and Average Workplace Quality,
Disaggregated, by Region
AU + NZ CIS EA ELAC MENA NA SA SEA SSA
Job Satisfaction -0.1188 -0.0528 -0.3898 0.0079 0.2699 0.197 8 0.1125 0.0502 0.2262 -0.0691
Workplace Quality
Pay
High Income -0.0211 0.3951 -0.2020 -0.0341 0.010 0 0.2882 0.00 16 0.2340 0.4419 0.0660
Individual Income
(Natural Log)
0.0678 0.3580 1.0677 -0.2534 0.0646 -0.1358 0.4904 -0.0899 -0.3507 -0.3231
Working Hours
Working Hours
(Natural Log)
-0.3226 0.1493 0.1481 -0.0145 0.0932 -0.0763 -0.0212 0.0978 -0.0223 0.1354
Working Hours Mismatch
Wants to Work
Same Hours
0,1627 -0,3300 -0,0142 0,0841 -0,4123 0,0123 0,070 1 -0,2590 -0,2936 -0,4503
Wants to Work
More Hours
-0.1728 0.4860 -0.0199 -0.0618 0.1636 0.0387 0.0981 0.3808 0.3691 0.6614
Wants to Work
Less Hours
0.0212 -0.1922 0.1627 -0.0189 -0.0899 0.0423 -0.0878 -0.1501 0.163 7 -0.1854
Work-Life Imbalance
Working on
Weekends
0.0455 -0.1851 0.3347 -0.0792 0.1093 -0.2606 0.1296 0.2328 0.4221 0.0685
Work Interfering
With Family
0.3016 -0.2337 -0.2183 0.0 150 -0.1286 -0.0809 0.1908 0.4872 0.1090 0.0517
Difficulty of
Taking Time Off
-0.2802 0.5383 0.0643 -0.0232 0.0227 0.0831 -0.2570 0.0132 0.1211 0.1069
Skills Match
Skills Match 0.2032 -0.3883 -0.2257 0.0848 -0.2645 -0.0169 0.2284 -0.4239 -0.3829 -0.2518
Skills Training 0.2063 -0.5247 -0.1103 0.0760 -0.1843 -0.1500 0.3068 -0.3587 -0.1740 -0.0558
Job Security -0.1380 -0.0051 -0.1868 0.0021 0.0740 0.1414 0.1963 0.1 974 0.1940 0.0146
Difficulty, Stress, Danger
Hard Physical
Work
-0.0265 -0.2031 0.1309 -0.0369 -0.0348 -0.2860 0.6011 0.6788 0.4902 0.2202
Stressful Work 0.0697 0.0562 0.0683 0.0 165 -0.2623 -0.1693 0.0789 0.2398 0.0181 -0.1176
Opportunities for Advancement
Opportunities for
Advancement
-0.0355 0.2047 -0.3084 -0.0598 0.2372 0.2177 0.3072 0.4015 0.5904 0.4213
Independence
Independent
Work
0.2337 -0.4169 -0.4156 0.0688 -0.0590 0.0113 0.2453 -0.2613 0.2347 -0.1349
Working From
Home
-0.0756 -0.1259 -0.0748 -0.1181 0.1305 0.2148 -0.0100 0.9287 1.2560 0.1375
Daily Work
Flexible
0,1280 -0,1832 -0,0541 0,0928 -0,3405 -0,1104 0,0991 -0,3108 -0,2932 -0,4651
Daily Work Fixed -0.1179 0.6073 -0.0106 -0.0264 0.2011 0.1123 -0.2818 -0.0415 -0.3789 0.5060
Daily Work Free -0.0098 -0.3662 -0.0104 -0.0662 0.1391 0.0192 0.2197 0.3151 0.7520 0.0176
Working Hours
Flexible
0,2356 -0,2920 -0,2064 0,0782 -0,2497 0,0756 0,147 2 -0,2896 -0,0920 -0,3727
Working Hours
Fixed
-0.1261 0.4643 0.0231 -0.0100 0.1291 -0.1873 - 0.0303 0.0025 -0.4687 0.4268
Working Hours
Free
-0.1162 -0.2496 0.1144 -0.0733 0.1088 0.1713 -0.1161 0.3999 0.8420 -0.0862
Working Schedule
Flexible
-0,0762 0,1464 -0,1103 -0,0068 0,0382 0,2062 0,0432 -0,0077 0,3422 -0,0291
Working Schedule
Fixed
0.1070 0.0187 0.0837 0.0138 -0.0681 -0.2497 0.1336 -0.2133 -0.5071 -0.0051

Global Happiness Policy Report 2018
AU + NZ CIS EA ELAC MENA NA SA SEA SSA
Working Schedule
Free
-0.0519 -0.0388 -0.1137 -0.0183 0.0831 0.03 13 -0.0478 0.2587 0.4254 0.1158
Interesting Job
Interesting Job -0.0 079 -0.1335 -0.5035 0.0634 0.0741 0.0673 0.1664 -0.4613 0.1549 -0.1434
Interpersonal Relationships
Contact With
Other People
0.1540 -0.3919 -0.1674 0.0595 0.0173 -0.0536 0.2553 -0.8281 -0.1678 -0.3284
Relationship With
Management
0.0060 -0.2032 -0.1031 -0.0311 0.1460 0.2665 0.0532 0.0987 0.0913 0.2633
Relationship With
Co-Workers
0.0580 -0.3813 -0.2724 0.0525 0.0583 0.2435 -0.0844 -0.2394 -0.2224 0.2284
Usefulness
Helping Other
People
0.157 5 -0.3394 -0.1683 -0.0384 0.1419 0.2406 0.4172 -0.2626 0.2819 -0.0508
Being Useful to
Society
-0.0672 -0.0692 -0.1257 -0.0231 0.167 1 0.0137 0.1882 -0.1840 0.1030 0.0061
Notes: All variables are standardized with mean zero and standard deviation one; negative values (marked in shades of red) indicate negative
deviations from the average value of the variable across countries, positive values (marked in shades of green) positive deviations. Observations
are weighted using country weights. The sample is restricted to all individuals who state that they are working and who report working hours
greater than zero. AU + NZ: Australia + New Zealand, CIS: Commonwealth of Independent States, EA: East Asia, E: Europe, LAC: Latin America and
Caribbean, MENA: Middle East and North Africa, NA: North America, SA: South Asia, SEA: South-East Asia, SSA: Sub-Saharan Africa.
Source: International Social Survey Program, Module on Work Orientations, Year 2015
Table W10 continued

124
125
Table W11: Summary Statistics of Variables in Section 4
Variable Mean Standard
Deviation
Minimum Maximum Number of
Observations
Remarks
Outcome
Job Satisfaction 5.314 1.167 1 7 27,732 “How satisfied are you in your main
job?” (1/8)
Controls
Age 43.225 12.897 15 95 27,732 -
Female 0.491 0.500 0 1 27,732 -
Partnered 0.584 0.493 0 1 27,732 -
Separated 0.020 0.139 0 1 27,732 -
Divorced 0.088 0.284 0 1 2 7,732 -
Widowed 0.025 0.157 0 1 27,732 -
Years of Education 13.315 3.943 058 27,732 -
Number of Individuals in
Household
3.234 1.735 125 27,732 -
Number of Children in Household 0.562 0.913 017 27,732 Number of children between school
age and 17 years of age
Number of Toddlers in Household 0.273 0.604 0 6 27,732 -
Union Member 0.237 0.426 0 1 27,732 -
Workplace Quality
Pay
High Income 2.822 1.101 1 5 27,732 “[…] how much [do] you agree or
disagree that [… your] income is
high?”: (1) “Strongly disagree” to (5)
“Strongly agree”, =4+5
Individual Income (Natural Log) 9.214 2.389 218 27,732 -
Working Hours
Working Hours (Natural Log) 3.636 0.452 0 5 27,732 -
Working Hours Mismatch
Wants to Work More Hours 0.329 0.470 0 1 27,732 And earn more money
Wants to Work Same Hours 0.536 0.499 0 1 27,732 And earn the same money
Wants to Work Less Hours 0.063 0.244 0 1 27,732 And earn less money
Work-Life Imbalance
Working on Weekends 2.858 1.365 1 5 27,732 “[…] how often does your job involve
working on weekends?”: (1) “Never”
to (5) “Always”, =4+5
Work Interfering With Family 2.344 1.102 1 5 2 7,732 “[…] how often do you work at home
during your usual working hours?”:
(1) “Never” to (5) “Always”, =4+5
Difficulty of Taking Time Off 2.250 1.064 1 4 27,732 “How difficult would it be for you to
take an hour or two off during
working hours […]?” (1) “Not at all
difficult” to (4) “Very difficult”, =3+4
Skills Match
Skills Match 2.800 1.016 1 4 27,732 “How much of your past work experi-
ence and/or job skills can you make
use of in your present job?” (1)
“Almost none” to (4) “Almost all”,
=3+4
Skills Training 0.434 0.496 0 1 27,732 “Over the past 12 months, have you
had any training to improve your job
skills either at the workplace or
somewhere else?” (0) “No” and (1)
“Yes”, =1

Global Happiness Policy Report 2018
Job Security
Job Security 3.7 76 1.105 1 5 2 7,732 “[…] how much [do] you agree or
disagree that [… your] job is
secure?”: (1) “Strongly disagree” to
(5) “Strongly agree”, =4+5
Difficulty, Stress, Danger
Hard Physical Work 2.698 1.335 1 5 2 7,732 “How often do you have to do hard
physical work?”: (1) “Never” to (5)
“Always”, =4+5
Stressful Work 3.176 1.069 1 5 27,732 “How often do you find your work
stressful?”: (1) “Never” to (5)
“Always”, =4+5
Opportunities for Advancement
Opportunities for Advancement 2.776 1.137 1 5 27,732 “[…] how much [do] you agree or
disagree that [… your] opportunities
for advancement are high?”: (1)
“Strongly disagree” to (5) “Strongly
agree”, =4+5
Independence
Independent Work 3.815 1.097 1 5 27,732 “[…] how much [do] you agree or
disagree that [… you] can work
independently?”: (1) “Strongly
disagree” to (5) “Strongly agree”,
=4+5
Working From Home 1.990 1.290 1 5 27,732 “[…] how often do you work at home
during your usual working hours?”:
(1) “Never” to (5) “Always”, =4+5
Daily Work Fixed 0.264 0. 441 0 1 27,732 “I am not free to decide how my
daily work is organized.”: (1) “Yes”
and (0) “No”
Daily Work Flexible 0.426 0.494 0 1 27,732 “I can decide how my daily work is
organized, with certain limits.”: (1)
“Yes” and (0) “No”
Daily Work Free 0.280 0.449 0 1 27,732 “I am free to decide how my daily
work is organized.”: (1) “Yes” and (0)
“No”
Working Hours Fixed 0.514 0.500 0 1 27,732 “Starting and finishing times are
decided by my employer and I
cannot change them on my own.”: (1)
“Yes” and (0) “No”
Working Hours Flexible 0.326 0.469 0 1 27,732 “I can decide the time I start and
finish work, with certain limits.”: (1)
“Yes” and (0) “No”
Working Hours Free 0.143 0.350 0 1 27,732 “I am entirely free to decide when I
start and finish work.”: (1) “Yes” and
(0) “No”
Working Schedule Fixed 0.692 0.462 0 1 27,732 “I have a regular schedule or shift
(daytime, evening, or night).”: (1)
“Yes” and (0) “No”
Working Schedule Flexible 0.15 3 0.360 0 1 27,732 “I have a schedule or shift which
regularly changes (for example, from
days to evening or to nights).”: (1)
“Yes” and (0) “No”
Working Schedule Free 0.079 0.270 0 1 27,732 “I have a schedule where daily
working times are decided at short
notice by my employer.”: (1) “Yes”
and (0) “No”
Interestingness
Interesting Job 3.834 1.000 1 5 2 7,732 “[…] how much [do] you agree or
disagree that [… your] job is
interesting?”: (1) “Strongly disagree”
to (5) “Strongly agree”, =4+5
Interpersonal Relationships
Contact With Other People 4.233 0.852 1 5 27,732 “[…] how much [do] you agree or
disagree that […, in your job, you]
have personal contact with other
people?”: (1) “Strongly disagree” to
(5) “Strongly agree”, =4+5
Table W11 continued

126
127
Relationship With Management 3.910 0.902 1 5 27,732 “[…] how would you describe
relations at your workplace between
management and employees?”: (1)
“Very bad” to (5) “Very good”, =4+5
Relationship With Co-Workers 4.187 0.757 1 5 27,732 “[…] how would you describe
relations at your workplace between
workmates/colleagues?”: (1) “Very
bad” to (5) “Very good”, =4+5
Usefulness
Helping Other People 3.884 1.003 1 5 27,732 “[…] how much [do] you agree or
disagree that […, in your job, you]
can help other people?”: (1) “Strongly
disagree” to (5) “Strongly agree”,
=4+5
Being Useful to Society 3.947 0.947 1 5 27,732 “[…] how much [do] you agree or
disagree that [… your] job is useful to
society?”: (1) “Strongly disagree” to
(5) “Strongly agree”, =4+5
Source: International Social Survey Program, Module on Work Orientations, Year 2015
Table W11 continued

Global Happiness Policy Report 2018
Table W12: List of Countries
Covered in Section 4
Australia & NZ
Australia
New Zealand
CIS
Russian Federation
East Asia
China
Japan
Taiwan
Europe
Austria
Belgium
Croatia
Czech Republic
Denmark
Estonia
Finland
France
Georgia
Germany
Great Britain
Hungary
Iceland
Latvia
Lithuania
Norway
Poland
Slovakia
Slovenia
Spain
Sweden
Switzerland
Latin America & Carib
Chile
Mexico
Suriname
Venezuela
Northern America
United States
South Asia
India
Southeast Asia
Philippines
Sub-Saharan Africa
South Africa
Source: International Social Survey Program, Module on Work
Orientations, Year 2015