International data survey analysis: Well-Being Report for Kaunas University of Technology December 2010

Introduction

This report is based on day 2 of a 2-day seminar at Kaunas University of Technology on 19th

December 2010. Day 2 was delivered by the centre for well-being, nef (the new economics

foundation), a think-and-do tank based in London striving for a new economics which can

deliver people’s well-being, sustainability and social justice. The seminar itself was entitled

“International data survey analysis”. The first day was delivered by Dr. Atle Jastad from the

Norwegian Data Archive (NSD), where participants were introduced to the European Social

Survey (ESS), the ISSP and Nesstar – an online tool for analyzing data.

Day 2, reported upon here, focused on the measurement of well-being in the context of the

search for alternative measures of progress. The day was combination of presentations and

activities for participants. The activities that actually took place differed slightly from those

planned. What is reported on here is a combination of the original presentation plans, and a

short description of the activities that actually took place.

nef gave three presentations as part of the day’s seminar:

1. Why measure well-being?

2. How to measure well-being

3. Beyond well-being?

Between presentation 2 and 3, participants spend time analysing data from the European Social

Survey using a combination of SPSS, Nesstar and Excel.

This report covers presentations 1 and 2 in detail, and briefly summarises presentation 3.

About nef

nef is an independent think-and-do tank founded in London in 1986 as the legacy organization

of The Other Economic Summit – an alternative meeting set up in parallel to the G7 to discuss

economic issues from the perspectives of people’s well-being, sustainability and social justice.

nef currently has around 40-50 staff on works on a range of topics including:

• well-being

• valuing what matters

• business and finance

• natural economies

• connected economies

• energy and climate change

• social justice

• democracy and participation

nef also co-ordinates the new Great Transition campaign which aims to bring together

campaigners working on a wide range of issues and highlight the common purpose and focus

efforts on the key changes that need to happen for a better world.

The centre for well-being was founded in 2006, though nef had been working on well-being for

several years before that. The centre’s aim is to enhance individual and collective well-being in

ways that are environmentally sustainable and socially just. We work to influence both policy

and the public – by changing the way we measure progress, by encouraging a questioning of

values, by highlighting the evidence around what does and doesn’t lead to high well-being, and

by exploring how policy-making could be different with a focus on well-being.

Why measure well-being?

The importance of measurement

Focussing on measurement may seem like a strange thing to do with all the grave problems

facing the world – climate change, resource limits, economic instability and growing inequality.

And yet we believe that how we measure progress may indeed be central to many of these

problems. Since the middle of the 20th century we have focused on one particular way to

measure progress – GDP (gross domestic product). Policies which are expected to increase

GDP are favoured over those that are not likely to, and policies that are expected to increase

GDP the most are favoured over those that might only increase it a little – usually disregarding

social and environmental impacts, and often with undue caution as to even the long-term

economic sustainability of the policies – as has been seen in the crash of 2008.

So, wealthy nations such as Canada exploit hard to reach tar sands so as to maximum growth,

with little regard for their greater contribution to climate change, and the harm caused to

ecosystems and communities living in the area.

Poor countries welcome

multinational companies in export

zones where they pay reduced

taxes and are able to pay minimum

wages and ignore labour laws. The

countries compete with one another

in cutting regulation so as to ensure

companies come to them and

bolster their weak economies.

Meanwhile, in places like Nigeria,

the government has been accused

of being complicit in some of the

harshest treatment of local

populations, again in the rush for

growth. Huge areas of forest are

cleared, to allow agricultural land

to grow crops to export to

wealthy countries to feed

excessive meat consumption far

beyond what can be considered

healthy, and only tolerated

because it too contributes to economic growth.

There are more subtle impacts in developed countries too. Why don’t governments take steps

to support reduced working hours when there is plenty of evidence that many people would be

willing to take a cut in their income and work less provided they did not feel it would harm their

careers? Why don’t governments seek to restrict obesogenic consumption and lifestyles given

the huge impacts on health and the huge costs to health care services? Why do governments

continue to allow the sale of arms to countries whose human records they simultaneously

condemn? Why do local authorities support, indeed encourage, the construction of out-of-town

superstores despite them ripping the souls out of communities and fostering car-dependent,

fuel-intensive lifestyles?

Each of these cases is of course complex and cannot be dealt with in a single paragraph.

However, it is clear that the quest for economic growth is a key factor in these decisions. GDP

growth is the trump card which can defeat other costs and benefits in policy decisions.

Governments strive for green growth, smart growth and socially just growth, but in few cases do

they ever consider letting go of that second word. So, for example, the EU’s 2004 Lisbon

strategy stresses that:

“the promotion of growth and employment in Europe is the next great European project"

Nowhere is this trump card played in a more deadly fashion than in relation to climate change.

Perhaps the single greatest reason that governments around the world have not effectively dealt

with climate change by limiting emissions has been that limiting emissions enough will inevitably

mean limiting growth and indeed perhaps even reversing it. For all the talk of ‘green growth’ and

‘green business’ the reality is that there is a conflict between continuing economic growth and

avoiding the worst of

climate change, as is

argued by the UK

Sustainable

Development

Commission report

Prosperity without

growth. And

continuing economic

growth is still winning

the battle.

Perhaps this is not surprising. As economist E.J. Mishan stated in 1967, amongst most

economists and politicians:

“…any doubt that, say, a four per cent growth rate is better for the nation than a three per

cent growth rate is near-heresy; is tantamount to a doubt that four is greater than three”

GDP has also captured the attention of the general public. Figure 1 shows the number of times

each keyword was cited in major world newspapers in two two-month time periods, according to

the online media search tool Nexis. GDP ranks top in both periods. One should not be deceived

by the apparently high number of articles citing well-being. Most of these are referring to well-

being in a very general sense – from discussions of woman’s health to yoga to tourism. Few are

talking about well-being as an indicator. Indeed most worrying is that the Human Development

Index, developed as a direct alternative to GDP by the UN, theoretically the most important

global institution, can barely be seen on this graph. In some ways it appears that, as suggested

by Jeroen van den Bergh, we are in ‘conceptual lock-in’

Figure 1: Number of citations of keywords related to measurement of progress in major world newspapers, according to Nexis

Figure 2: Importance of seven concepts rated by students in 47 countries around the world. Source: Diener & Scollon (2003)

0

500

1000

1500

2000

2500

3000

GDP

Unemplo

yment

rate

Life e

xpec

tancy

Well-bein

g

CO2 emiss

ions

Ecologic

al foo

tprint

Human Dev

elopm

ent In

dex

Num

ber o

f arti

cles

Jul & Aug 2007Jan & Feb 2008

1

2

3

4

5

6

7

8

9

Happin

ess

Health

Love

Meanin

g

Wea

lth

Heave

n

Attrac

tiven

ess

Impo

rtan

ce

More than economics

Having said that, at least some part of us recognises that growth is not the answer to

everything. In an international survey 75% reported believing that environmental, health and

social indicators should be given as much weight as economic ones. Meanwhile, a poll

conducted for the BBC in Britain found that 81% of respondents thought government’s prime

objective should be to work for ‘greatest happiness’ rather than ‘greatest wealth’. In a 2003

study, Diener and Scollon asked people (admittedly all students) around the world in 27 nations

from the USA to Uganda what they most felt to be important to them (see Figure 2). Happiness

ranked top, followed by health, love and meaning. Wealth is only in fifth place out of the seven

possibilities given in the study.

For most people, therefore, there is something that is more important than wealth as purportedly

measured by GDP (of course even the link between GDP and the wealth of most individuals is

tenuous, as has been highlighted by the Stiglitz Commission amongst others). As the early

psychologist William James wrote:

“How to gain, how to keep, how to recover happiness, is for most people, at all times ...

the secret motive of all they do, and of all they are willing to endure”

Given that this happiness is more important, isn’t it useful to know how each country is doing in

terms of achieving it? Might we not learn from such information how to better achieve it?

Government’s role in well-being

Some have argued that well-being or happiness are personal matters and it should not be

government’s role to meddle in it. But government does seek to improve health, which is also a

personal matter. There is good reason for this. Government should be a tool for the population,

helping us to achieve things that are hard to achieve on our own. In relation to well-being this

can work in five ways, none of which are new to political science.

Firstly, it can help those with few resources to achieve well-being. This has an intrinsic value for

those being helped, improving their well-being. But it also has instrumental value, in that it can

reduce the risks of social tension that are linked to inequality and dissatisfaction amongst

population groups.

Secondly, it can set rules which protect one group or person from harm caused by another

group or person. Our system of justice aims to do this. For example we prohibit theft as we

believe that it causes unfair harm. Might some other actions, which clearly harm well-being,

also be subject to law?

Thirdly, it can resolve social action problems. Sometimes, the rational behaviour of an individual

can have unintentional negative impacts on the well-being of others. So, for example, an

individual’s choice to drive might mean they can travel faster, but the resulting traffic means

that, overall, the population travels slower. Some governments, local and national, therefore

attempt to manage traffic to reduce such social action problems.

Fourthly, it can think about the future in ways that individuals might not be motivated to do. This

has been a particular challenge for most countries, but efforts to curb climate change by

reducing emissions at the moment are an attempt to think about the well-being of future

generations. Without governments, these efforts would be very difficult.

Lastly, actions by government to improve well-being can be more efficient than those of an

individual. It is more efficient for communities to pool together their resources and build a library,

than for each to buy and own all the books they need. It is more efficient for governments to

invest in information campaigns to improve well-being than for this to be done by word of mouth.

In other words, government has a legitimate role in supporting, protecting and sustaining well-

being within any political philosophy. With this in mind, it is only natural that it should seek to

measure its impacts on well-being.

What is well-being?

All the top four concepts in Figure 2 fit within our understanding of well-being. For us, it is

important to extend well-being beyond just happiness and/or satisfaction. These sensations are

part of well-being, but can be augmented by other experiences which are universally important

to people – such as feeling close to others, feeling autonomous and authentic, and feeling able

to contribute to something or someone. These experiences are more about ‘doing’ well, than

simply feeling good, and they contribute to sustaining and supporting the more general concepts

of happiness and satisfaction.

Our conception of well-being is presented in Figure 3, reproduced from the Foresight Mental

Capital and Well-Being Project, a government sponsored project commissioned to explore the

implications of well-being science and mental health issues. Good feelings are included in this

model, in the top oval. These are determined by good functioning and the satisfaction of the

needs noted earlier. The list highlighted here (being autonomous, competent, safe and secure,

connected to others) comes from self-determination theory and its development by psychologist

Tim Kasser. This in turn is determined by the external conditions an individual finds oneself in,

and the psychological resources they bring to their situation.

Figure 3: Dynamic model of well-being1

Well-being is dynamic, each aspect affecting the others – the arrows shown in Figure 3 highlight the major pathways in the model. This means that, no matter which aspect is seen as most important, all should be considered in terms of promoting well-being.

This understanding was recognised in the first UK government definition of well-being as:

“… a dynamic state, in which the individual is able to develop their potential, work

productively and creatively, build strong and positive relationships with others, and

contribute to their community. It is enhanced when an individual is able to fulfil their

personal and social goals, and achieve a sense of purpose in society”

1 Foresight Mental Capital and Well-Being Project (2008) Final project report (London: The Government Office for Science), pg.63

Measuring well-being

The following section of the seminar explored how well-being can be measured. First, participants went through a mock questionnaire, based on the European Social Survey well-being module, to get a flavour of how well-being is measured. Then, some of the issues with measuring well-being and with questionnaire design in general were considered.

Decisions in well-being measurement

Our understanding of well-being leads to the following implications in terms of measurement.

Most importantly, given our definition of well-being as being about how well people feel and how

well they are meeting psychological needs, it is clear that measuring it must be, at least in part,

through asking people how they feel, i.e. subjectively. Feelings are subjective experiences

and they are best assessed subjectively. No one can be a better judge of an individual’s

feelings, than that person themself. Of course, we can, based on what we know about

psychology and human biology, measure things which we know to be universally important to

people in determining their well-being through objective measures – for example calorie intake,

and social interaction. But, the well-being itself is something which happens within the individual

and therefore cannot be measured accurately through external observation.

Secondly, as well-being is multi-dimensional, no single indicator can fully capture it. At least one

indicator is needed for each hypothesised dimension. As such, a key question in measuring

well-being is what are the dimensions? Can the dimensions be prised apart such that one

aspect of well-being increases whilst another doesn’t?

Thirdly, measurement should consider whether there is any ‘structure’ for well-being. Are some

aspects more important than others? How should this be reflected in measurement – more

questions? Unequal weighting? Are some aspects of well-being negotiable, but others not?

Measurement options can reflect such decisions.

Principles of psychometrics

As well as the implications specific to measuring well-being, there are also some considerations

which apply to psychometric measurement in general. Many of these are framed here as the

ideal situation – in reality measurement cannot always follow these principles.

Firstly, psychometricians stress the importance of using multiple items in questionnaires to

measure a single construct. It is important to recognise that questions do not perfectly measure

what they purport to measure. Asking someone how satisfied with life they are on a scale of 0-

10, their answer is likely to be influenced by how happy they are, but also whether they like the

interviewer, how they answered the last question, what their favourite number is, what particular

aspect of their life they have been thinking about recently, whether they are prone to agreeing or

disagreeing, complaining or presenting a positive perspective, etc. Each question and each

response is therefore affected by a range of factors. By answering a set of questions, one hopes

that the idiosyncracies of response to one are smoother over. Each question will be affected by

its own idiosyncracies, but the intention is that the common variance between them represents

something fundamental – the construct one is trying to measure. So, for example, Rosenberg’s

Self-Esteem scale is calculated from 20 questions on the single construct of self-esteem.

When constructing questions, questionnaire-designers attempt to make sure they are simple

and unambiguous. It is better not to load too many concepts into a single question. For

example, asking “Is time spent with your family stressful or enjoyable?” may be problematic

because one has to think about two issues – how much of the time is enjoyable, and also how

much of the time is stressful. In the European Social Survey, two separate questions are asked.

It is important to use language that the vast majority of respondents will understand. It is also

important to only ask about things which will be relevant to as many people as possible. For

example, quality of healthcare or education services within an area may be important to some

people, but not to everyone. As such, it would probably be imprudent to ask too many questions

on such a topic.

A more subtle consideration is question order. The response to a given question is affected by

the questions that came before it. In an experiment conducted with students, two groups were

given subtly different questionnaires. In group A students were asked to rate their satisfaction

with life on a scale from 0-10, and then were asked to report on how often they went on dates.

In group B, this question order was reversed – first they were asked about how often they went

on dates, then they were asked their life satisfaction.

For group A there was no significant correlation between the two questions mentioned. For

group B, a significant correlation of r=0.66 was found. Before the questionnaire, the two groups

were identical, so we can assume there is no difference in the importance of dating to

individuals in the groups. Rather, it appears that being asked about their dating experience,

students then bore dating in mind more when considering their satisfaction with life. An

individual who was not satisfied with their love life might nevertheless report being satisfied with

life in general in group A. However, if they were in group B, and reminded of their poor love life

first, they might then report not being satisfied with life in general. This is called a focussing

bias. Generally, focussing biases are more problematic for more general questions such as life

satisfaction. For example, it does not appear that being asked to judge one’s life satisfaction

before the dating question would have affected responses to the latter question. As such, it is

recommendable to ask more general questions such as life satisfaction first, and then more

specific questions – a recommendation with might at first seem counter-intuitive.

As with all measurement, reliability and validity need also be considered when developing

metrics. Reliability can be tested using test-retest: how stable are responses to questions if one

were to ask the same questions twice with an interval in between. Of course, such an approach

must be treated with caution. Some questions are more likely to be unstable for good reasons

and not because of low reliability. For example, an individual’s true levels of self-esteem are

unlikely to change dramatically over a week, unless something life-changing were to take place.

On the other hand, asking someone how happy they have felt over the last day, their responses

are likely to change greatly from day to day and this need not be considered as error.

Validity has many facets. Firstly measures should have face validity. There should be some

reasonable logical interpretation of how they measure what they purport to measure. Secondly,

they should have criterion validity. This means that the measure should correlate highly with

other measures that purport to measure the same thing. For example, in the case of subjective

well-being, we know that an individual’s assessment of their happiness correlates with other

people’s assessments of their happiness, how often they smile, and neural patterns known to be

associated with positive emotions.

Thirdly, measures should correlate with factors believed to have an effect on them or to be

affected by them (convergent validity). For example standard measures of life satisfaction

correlates with income, marital status, social relations and a wide range of other variables which

one would expect to be important in determining how someone feels about their life.

Lastly, measures should not correlate too much with measures that purport to measure other

concepts, if those concepts are believed to be distinct. For example, self-esteem and feeling

satisfied with one’s relations may be correlated to some extent, but if measures of the two are

intended to be measuring two different things, then they should only correlate to some extent –

this is known as discriminant validity.

Real world compromises in measuring well-being

The principles of psychometrics have been developed by academics rightly intent on identifying

the best possible ways to measure psychological phenomena. Often they are most appropriate

for academic studies, where the academic has the most control over the measures being

deployed.

Proponents of well-being measurement (such as Professor Ed Diener, Lord Richard Layard,

Professor Ruut Veenhoven and ourselves at nef) have argued that government should collect

data on well-being from the general population to inform policy-maklng. Generally, it is

suggested that measures of well-being are included in official national statistics collection

processes.

Of course, in such a context, well-being measurement experts must co-operate with and

compromise with policy makers and, most importantly, official statisticians. These partners will

have very valid concerns about adding too many questions to surveys – they increase costs and

respondent-fatigue, and may not be seen as appropriate by respondents. These concerns mean

that, for example, the aim of including multiple-item scales to measure each aspect of well-being

may be a little optimistic. Indeed, it is even a challenge to convince statisticians to consider the

multi-dimensionality of a concept which, until recently, was unknown to them (at least in a

professional context!).

Indeed legitimate concerns about space often go hand-in-hand with concerns that we would

argue are less-justified with regards to the validity, reliability and usefulness of well-being

measurement. Statisticians that have focussed on asking more ‘factual’ questions regarding

individual income and habits, have little training in psychometrics, the recommendations it

makes with regarding measuring psychological constructs and the assurance it provides as to

their validity and reliability if done well.

Other problems of inertia are also encountered. National statisticians prefer to use questions

with a proven track record – but of course it is hard for such a track record to develop without

financial support from somewhere. Those in charge of specific surveys can also be rather

protective about changes that are to be made – not wishing to remove questions that have been

used before, or change question order. These factors can be problematic given what we have

already discussed about psychometric principles.

(life sat)

National Accounts of Well-Being

In this section we present a first

attempt at an approach that could be

used by national, regional or local

governments to measure well-being in

their jurisdiction. It is an approach that

builds upon theories of well-being, and

principles of psychometrics, attempts to

give consideration to the realities of

policy-making, and also the realities of

national statistics collection. As such it

represents a compromise to some

extent. It is also worth noting that our

theories of well-being have come a

long way since the approach began to

be developed in 2004 and, in years to

come, we expect to be able to update

it.

The National Accounts of Well-Being were released as a report2 and a website

(www.nationalaccountsofwellbeing.org) by nef in 2009. It was based on data from a module on

well-being included in the third round of the European Social Survey (ESS) in 2006. The 55 item

module was developed by a team lead by Felicia Huppert at the University of Cambridge, and

including Andrew Clark at the ENS in Paris, Bruno Frey at the University of Zurich, Johannes

Siegrist at Dusseldorf University, and Nic Marks at nef. The ESS is a biannual survey and is

designed such that a core set of questions is asked each wave, accompanied by two ‘rotating’

modules that are developed by external academics.

In 2010, a new team including Felicia Huppert and ourselves won the opportunity to develop a

repeat well-being module for the sixth round of the ESS in 2012, which is work we are now in

progress with. Our aim is to improve the module we developed for ESS3, though of course, as

with any pre-existing survey vehicle, we must consider constraints such as preserving 60% of

the questions from round 3, and, of course, having no say about questions included in the core

of the survey.

In 2006, the ESS reached around 40,000 respondents in 22 European countries. More recent

waves have managed to reach more countries within Europe.

Structure for measuring well-being

One of the key innovations of the National Accounts of Well-Being was to develop a structure

for measuring well-being. Over 50-items were available for assessing well-being. How would

these be brought together to produce a coherent and useful approach?

Figure 4, reproduced from the report, is our solution. The questions were brought together in a

set of constructs such as positive feelings, satisfying life or resilience. Many of these concepts,

in turn, formed larger constructs such as functioning. All together, the questions were brought

together as either personal well-being and social well-being. In reality, in the report, scores are

also calculated combining these two high-level concepts.

The idea is that one can look at well-being at various levels of aggregation. A headline figure

can be produced to assess well-being in general But one can also look at specific aspects of

2 Michaelson J, Abdallah S, Steuer N, Thompson S & Marks N (2009) National Accounts of Well-being: Bringing real wealth onto the balance sheet (London: nef)

well-being, and explore how a given population group or country is doing in terms of self-

esteem, for example, or social relations – allowing priorities and problem areas to be identified.

Figure 4: Indicator structure within the example national accounts framework

Figures 5 to 8 show how the data from the National Accounts can be used. Figures 5 and 6

show the overall personal and social well-being scores respectively for all the countries in the

report, as presented on the website. Shades of green represent high well-being and shades of

red represent low well-being.

Figure 5: Personal well-being

Figure 6: Social well-being

Figures 7 and 8 show how the data for several dimensions can be combined in one diagram,

creating a ‘well-being profile’ for a country or, in the case of Figure 8, a subset of the population

in a given country.

Figures 7 & 8: Well-being profiles for Spain and UK

3,004,005,006,007,00+ve feelings

-ve feelings

satisfying life

vitality

resilience & self-esteem

postiive functioning

supportive relationship

s

trust & belonging

Spain vs. UK

3,004,005,006,007,00+ve feelings

-ve feelings

satisfying life

vitality

resilience & self-esteem

postiive functioning

supportive relationships

trust & belonging

Low income groups

SpainUKEuropean average

Calculation metrics

How are questions brought together to calculate scores at the various levels? Figures 7 and 8

show scores on a scale of 0-10. Importantly, how can scores in different components of well-

being be compared? This is not a simple task. One cannot simply take an average of answers

to each question, because they are based on different scales, some from 0-10, some simply yes

or no. Some questions are worded such that a high number means high well-being, for others it

is the reverse. In any case, even if the wordings and scales were the same, one could not be

sure that a 7 for one question actually represents higher well-being than a 6 on another.

Consider the following two statements assessed on the same scale:

“I’m always optimistic about my future”

“There are people in my life who care about me”

The former question is designed to assess optimism, the latter assesses close social relations.

Both are answered on a 5-point scale from strongly disagree to strongly agree. In both cases,

the highest well-being is to respond ‘strongly agree’. But there the similarities stop. Far more

people agree strongly with the second statement than the former. Does that mean that overall

people’s social relations are better than their optimism? Maybe not. Consider if the first

question was changed slightly to “I’m usually optimistic about my future”. In that case one

would expect that many more people would agree with the statement, perhaps even more than

the social relations question. The fact that the wordings of the questions can have a big impact

on the responses means we cannot compare mean scores for two questions directly and hope

for anything meaningful. Indeed, it would be very hard to produce any results that could allow us

to say, in absolute terms, that people in Europe have greater optimism, than, positive relations.

The best we can hope for is to compare sub-groups within Europe, relative to the mean. We can

then say that people in Spain, for example report better supportive relationships than sense of

vitality, compared to the European average. What this means is that Spanish people have

above average responses in terms of supportive relationships, but only at average for sense of

vitality.

To determine these patterns, to produce scores from 0-10, we developed a three stage process.

Firstly, we used a standard statistical technique to standardize results for each question, as

shown in the equation below.

responsesallfordeviationstandardtheisresponsesallofmeantheisx

xxscorez

σ

σ−

=.r

The formula results in a z-score for each individual for each question, which is basically the

distance away from the mean for that question across all Europe, in units of standard deviation.

If someone scores above average on a question, they will have a positive z-score. If they score

below average, they will have a negative z-score. The units can be meaningfully compared

between two questions. If someone scores 2.00 on one question, and 1.00 on another, that

means that they are further above the mean on the first one – and so one can conclude they

having higher well-being in that aspect.

The next step is to aggregate results to calculate components scores. This is simply a matter of

taking averages,

The last step is more unusual. Z-scores may be okay for academics, but they are not very

meaningful or understandable for anyone else. Importantly, one does not have a sense of what

the maximum or minimum is. To deal with this, we developed an equation to transform z-scores

into scores on a scale of 0-10 such that 0 represents the lowest possible score, 10 the highest

and 5 the mean (i.e. 0 as a z-score). The function is demonstrated in the graph in Figure 9 for

an example question:

Figure 9: Transformation function from z-score to 0-10 scale

0123456789

10

-3 -2 -1 0 1

z-scores

trans

form

ed s

core

s

The result is neat 0-10 scores which can be calculated at any level of aggregation, be it a single question, or well-being overall for a nation.

Beyond well-being?

After participants spent several hours exploring well-being data, the final presentation explored

the measurement of progress beyond well-being. Of course well-being is important but there is

more to progress. Figure 10 presents a prototype measurement framework.

Figure 10: Measurement framework

The framework highlights three spheres of measurement – human well-being, the environment

and other human systems, particularly the economic system. It also highlights the interactions

between them as being important to study. So, for example, decoupling between economic

activity and environmental impact is desirable. Similarly the efficiency of well-being output per

unit of economic activity should be considered – how efficient is our economy in achieving our

goals. Most important, based on this framework, however is the relationship between the

environment and human well-being. In a way, this is the ultimate efficiency measure – how

much well-being is achieved per unit of environmental impact.

Happy Planet Index

The Happy Planet Index, developed by nef in 2006, is an attempt to operationalise the

ecological efficiency.3 The Happy

Planet Index takes life satisfaction

(a measure of subjective well-

being) and life expectancy (a

measure of the key health

outcome) and multiplies them

together. The resulting figure is

called happy life years, based on

the original approach developed by

Dutch sociologist Ruut Veenhoven.

It then divides this figure by the

ecological footprint of a country (the

land required to produce the

resources and sequester the CO2

produced by a nation, based on

consumption patterns). A

couple of adjustments are

made to control for the

differences in variation in the

different data sources.

The results are quite unique, but makes sense when one considers what is being calculated. As

is shown in Figure 11, the desired situation is of course high happy life years and low ecological

footprint. But most countries have one or the other. Rich Western nations have happy life years

and a large footprint. Poor African nations have small footprints, but low happy life years.

However, the graph also shows the relationship is not generally linear. There appears to be a

case of diminishing returns with increasing footprint contributing more to well-being for poorer

countries. Also, of course, there is much variation between countries, with Latin American

countries appearing to be more ‘efficient’ in converting their middle-sized footprints into well-

being than other countries with similar sized footprints, such as the former communist nations of

33 Abdallah S, Thompson S, Michaelson J, Marks N & Steuer N (2009) The (un)Happy Planet Index 2.0; Why good lives don’t have to cost the Earth (London: nef)

Central/Eastern Europe and Central Asia. Figure 12, shows what happens when combines

these figures in a single map. Green indicates a greater approximation to ecological efficiency,

red less.

Figure 11: Happy life years and ecological footprint

As can be seen in Figure 12, Latin American countries fare the best, with Costa Rica taking top

position in the HPI scoring (see Figure 13)

Figure 12: Happy Planet Index coloured map

HPI rank Countries Region Life Sat Life Exp EF HPI

1 Costa Rica 1a 8.5 78.5 2.3 = 76.1

2

Dominican Rep 1a 7.6 71.5 1.5 = 71.8

3 Jamaica 1a 6.7 72.2 1.1 = 70.1

4 Guatemala 1a 7.4 69.7 1.5 = 68.4

5 Vietnam 6c 6.5 73.7 1.3 = 66.5

6 Colombia 1b 7.3 72.3 1.8 = 66.1

7 Cuba 1a 6.7 77.7 1.8 = 65.7

8 El Salvador 1a 6.7 71.3 1.6 = 61.5

9 Brazil 1b 7.6 71.7 2.4 = 61.0

10 Honduras 1a 7.0 69.4 1.8 = 61.0

20 China 6a 6.7 72.5 2.1 = 57.1

35 India 5a 5.5 63.7 0.9 = 53.0

51 Germany 2c 7.2 79.1 4.2 = 48.1

74 UK 2c 7.4 79.0 5.3 = 43.3

114 USA 2b 7.9 77.9 9.4 = 30.7

143 Zimbabwe 4a 2.8 40.9 1.1 = 16.6

Figure 13: HPI scores and ranks for selected countries

Gathering momentum

This final section summarises some of the other developments in the field of measuring

progress taking place around the world.

Perhaps most influential has been the so-called Stiglitz Commission (officially the ‘Commission

on the Measurement of Economic Performance and Social Progress’) set up by French

President Nicholas Sarkozy in January 2008. The Commission, headed by Joseph Stiglitz,

Amartya Sen and Jean-Paul Fittousi and consisting of many other leaders and experts in the

fields of well-being, environmental assessment and economics, resulted in 21 recommendations

divided into three domains: quality of life, the environment and ‘classical GDP issues’ – the latter

referring to how to improve or supplement GDP in measuring economic issues. Amongst the

recommendations made were to develop and integrate subjective measures of well-being in

national statistics.

Before the Stiglitz Commission had been set up, however, the OECD had already been working

on measuring progress. Their work kicked off with the Istanbul Declaration in the Autumn of

2007. Later they produced a framework for measuring progress, which, like ours, highlights the

importance of well-being and the ecosystem. In 2011, the OECD will be launching a Handbook

of Subjective Well-being with clear guidelines of National Statistics Offices on measuring

subjective well-being.

Figure 14: OECD Framework for Measuring Progress

The polling company Gallup have also been actively involved. They have been collecting data

on well-being in the US from 1000 individuals every day. They have also launched the Gallup

World Poll which covers well-being to some degree.

Within the European Commission, the Beyond GDP process began with a seminal conference

in the European Parliament in November 2007. They have a road map for incorporating

alternative indicators and in 2008 a feasibility study for measuring well-being for Europe began,

a second study beginning in 2011.

Lastly, many countries are getting actively involved. In the UK, David Cameron recently

announced that the UK Office of National Statistics would be collecting well-being data and that

it was time to acknowledge that there was more to national progress than GDP.