Understanding happiness: ways to measure wellbeing

As well as the national headline measures of personal wellbeing, new methods of understanding ‘how we are doing’ are being used and studied. Here is a spotlight on the some recent findings.  We are exploring more of these methods in our measuring wellbeing series.

Measuring happiness from the words we use

Happiness levels of citizens of the UK 1771-2009

A recent study by The Centre for Competitive Advantage at Warwick University and the think tank the Social Market Foundation looked used the emotion words in over eight million books to map happiness over 200 years in six countries. They found that tracking the words used gave very similar findings to more traditional measures of wellbeing.  

As you might expect, war, civil conflict and economic collapse is very bad for our wellbeing and increase life expectancy and decreased child mortality increases happiness. They concluded we were happiest in 1957. This might come as a surprise to many people of colour, the LGBT community and other minorities who lived during the 50s; obviously, what is published in books is not the whole story.

We can see the impact on our happiness of more recent national events from the Office for National Statistics data here in the UK.  Grounds for an autumn bank holiday perhaps?


Taking the pulse of the public mood online

Other teams are now looking at how the words we use online tell us about population level happiness and predictors.

#HappySheffield shows the last 50 tweets analysed from Sheffield and the lastest emotion in the city.  You can compare your own twitter activity.


Dow Jones Index of Happiness who find that, unlike the age findings below, we are most unhappy as teenagers 13-14 years, with happiness rising to 45-60 and then down again for 75-84 years.


The World Wellbeing Project Has a range of studies on the language we use and our wellbeing including a study that was able to predict county level incidence of atherosclerotic heart disease from the use of words reflecting negative social relationships, disengagement, and negative emotions—especially anger—emerged as risk factors; positive emotions and psychological engagement emerged as protective factors.

Other happiness nuggets

The report has a lot of nuggets of recent wellbeing research findings

  • Happier workers are 12% more productive and that traumatic life events e.g. death or serious illness of close family member, unsurprisingly, reduces happiness but also reduces productivity.
  • Wellbeing research can explain voting behaviour beyond traditional financial indicators with drops in life satisfaction leading 10-12% drop in support for the governing party.
  • Happiest places have fewer unhappy people rather than having extraordinarily happy people but happiest places tend to have higher suicide rates.
  • Genetics explains up to ⅓ of our wellbeing with some people being more sensitive to their environment and the impact of both positive and negative life events.  
  • Happiness is contagious  
  • Happiness is U-shaped aka the midlife crisis is real


  • Eat your SEVEN a day – Eating fruit and veg increases mental wellbeing even more, and more quickly, than it improves physical wellbeing.


Do this

We can target happiness gains when decision making.  Policy makers already use wellbeing data to inform their case to HM Treasury for societal impact and the refreshed Green Book will give greater prominence to wellbeing metrics. Understanding of what makes us happy can help direct limited resources towards most effective ways to improve happiness.  

The report concludes with the policy recommendations listed below.

  • Vibrant economy and economic stability are important for wellbeing not just income growth. Stable employment and avoiding runaway inflation as the aim avoids the very negative impacts on wellbeing of unemployment beyond loss of income and allows spending on other things that enhance wellbeing.
  • Mental health services that are affordable, widely available, easily accessible and less stigmatised.
  • Better health including efforts to increase longevity, reduce child mortality and cleaner air including focus on prevention.
  • Shorter, and I would argue, better quality commutes.
  • Look beyond individual impact by fostering strong social networks to combat loneliness and because happiness is contagious.
  • Research which feelings matter most.

Lifelong learning: what are the wellbeing co-benefits?

The Government’s recently published Industrial Strategy, now out to consultation, indicated a shift towards embedding lifelong learning to help people retrain in new skills and adapt to rapid changes in technology.  What the strategy misses is how learning directly, and indirectly, affects our wellbeing.

There are 1.9 million adults enrolled in further education colleges, according to the Association of Colleges. And thousands of employees in the UK participate in work-based training and development courses each year.

When we spoke with the people around the UK about wellbeing, the most commonly used word was ‘opportunity’.  What you told us about learning, work and wellbeing was that:

  • work and learning offers achievement, satisfaction, appreciation, pride
  • work and learning can bring a sense of fulfillment, belonging, shared interest and experience
  • transition points (such as going into and out of retirement) need support.

What’s more, learning is also one of the evidence-based Five Ways to Wellbeing along with give, connect, take notice and be active.

And the obvious goal of the countless courses and training on offer appear to be improving they way we work, or changing our work completely.

But the evidence reveals that learning has an interesting and complex relationship with wellbeing. In the short term, the impact on wellbeing can be negative. Meeting essay deadlines or taking exams, for example, can increase stress or reduce social interactions with friends, which are factors that contribute to our subjective wellbeing.

When you take a longer view, however, additional formal qualifications or education shows a slightly positive impact on a person’s wellbeing. Again, this seems to correlate to what we might expect. For instance, we know there is a strong link between employment, especially high-quality jobs, and wellbeing. So where qualifications can lead to employment and higher quality jobs, it appears to be a good thing.

It also makes sense that where the process itself of formal and informal learning reduces isolation, there are benefits to participants’ wellbeing.  

But what, if any, are the benefits of lifelong learning to us as individuals, beyond its impact on employment progression? Does it matter what we learn about, or if we learn in-person or online? What are the criteria that make learning meaningful? And how does learning change over time? One striking finding about learning in the UK, for example, is that it is the one of the five ways to wellbeing that drops off most dramatically with age.

Evidence from observational studies (non-intervention studies that assess naturally occurring levels of learning, such as qualification or accreditation gained, or job status) supports a relationship between wellbeing and learning. Although identifying the causal mechanism has been more difficult: do happier people engage in more learning opportunities, or do learning opportunities make people happier?

In February, the Centre will publish our findings from a systematic review carried out with our team in the University of East Anglia on the wellbeing impacts of learning at work. Later in the year, we will publish the results of a systematic review of adult learning and wellbeing. This explores the processes through which this positive wellbeing occurs. 

Understanding the potential health benefits of different types of learning obviously has implications for service delivery, funding and policy making in further education and workplace learning.

The findings will address the question about the benefits of lifelong learning beyond employment-related benefits. It’s an important one. Especially if we are to evolve our perception of work-based and community learning from being focused mainly on what it does for our employment prospects, to how it contributes to our personal and community wellbeing. 

To get an alert when the work and learning findings become available, please email us at info@whatworkswellbeing.org.

Five ways to Wellbeing in the UK

5The Five Ways to Wellbeing are a set of evidence-based actions which promote people’s wellbeing. Whilst not claiming to be the biggest determinants of wellbeing, it’s a set of simple things individuals can do in their everyday lives. They were developed by the New Economics Foundation and based on the findings of the 2008 Government Office for Science Foresight report on Mental Capital and Wellbeing that aimed to develop a long term vision for maximising wellbeing in the UK.

They are

  1. Connect
  2. Be Active
  3. Take Notice
  4. Keep Learning
  5. Give

The 5 ways to wellbeing are integral to many activities that we care about and enjoy. Since their publication, the five ways have had an enormous reach, being used as evaluation frameworks, in school curriculums and by local authorities . They have formed the basis to specific interventions to improve wellbeing that we will be reviewing as part of our community wellbeing programme. 

In 2012, the European Social Survey was the first major survey to include questions directly on the five ways to wellbeing, allowing exploration of patterns of five ways behaviours across Europe for the first time.

fig19The Making Wellbeing Count for Policy research by Cambridge University, City University and the New Economics Foundation looked at this rich survey data and found:

  • People in the UK have low levels of participation in the five ways to wellbeing, compared to peer countries such as France and Germany particularly on Take Notice. With the exception of those aged 65 and over.
  • Having children seems to limit people’s opportunities to take notice in the UK in ways that do not apply in the rest of Europe.  
  • People of working age in the UK connect less than their peers in the rest of Europe, though this deficit also applies to those not in employment, suggesting that it cannot be explained purely in terms of working patterns.
  • Those in the UK aged 25-64 were much less likely to connect than their peers in other countries
  • Young women (15 – 24), parents, and people doing housework or childcare in the UK reported very low rates on Take Notice (whether people take notice and appreciate their surroundings). This finding was not replicated across Europe, suggesting there may be particular barriers in the UK for these population groups which may be amenable to policy.

The figures show the UK’s levels of five ways participation ‘Connect’ compared to other countries. Countries with a GDP per capita of below $30,000 are shaded in lighter blue. Given that the UK has a GDP per capita of almost $40,000 one would expect it to achieve higher participation in five ways than those countries.

Previous research has found that generally, in the UK:  

  • Males are more likely to be active, whereas females are more likely to give and connect.
  • People from lower socio-economic groups are less likely to be active, give and keep learning.
  • The older people are, the less likely they are to keep learning and be active.
  • However, for both Connect and Give, the trend follows a U curve, with people aged 16-25 and 65-74 most likely to engage in these activities.
  • People with qualifications are more likely to keep learning and give.

First quarterly UK personal wellbeing update to September 2016


Fig 1b average anxiety ratings to Sept 2016

Fig 1b average anxiety ratings to Sept 2016

Average ratings of anxiety increased slightly between the years ending September 2015 and 2016.


Wales was the only country to have higher anxiety ratings than the UK average.




Figure 1a Average life satisfaction, worthwhile and happiness ratings to Sept 2016

Figure 1a Average life satisfaction, worthwhile and happiness ratings to Sept 2016


Average life satisfaction, worthwhile and happiness ratings were unchanged between the years ending September 2015 and 2016.






New rolling quarterly updates

Measuring personal well-being can help us understand how changes in circumstance, policies and wider events in society may affect people’s lives, perhaps more so than traditional economic measures. Up to now, the current publication of official well-being statistics by the Office for National Statistics only allowed for changes in personal well-being to be assessed once a year.  With growing demand for timely statistics ONS reduced the lag from the end of the reporting period to 4 months from 6 months. And we look forward to more improvements to the timeliness of these statistics.

In a bid to make the personal wellbeing data more usable by the public and policy makers alike, ONS are publishing for the first time today, annual UK and country estimates for the 4 personal wellbeing questions on a rolling quarterly basis. Personal well-being data will, going forward, be published 4 times a year. Today’s publication contains data from year ending March 2012 to year ending September 2016.

Housing and wellbeing: special focus

To kick off 2017 we are publishing our scoping review on housing and wellbeing. It looks at what we know and where the gaps are. It also sets out what our focus will be for our upcoming systematic review on housing later this year. We’re focusing on housing and wellbeing because that’s what emerged as a priority area for a range of stakeholders in our public dialogues.


Read the full scoping review or the four-page briefing.



What does the English Housing Survey tell us about housing and wellbeing?

This week the What Works Centre for Wellbeing shares a new scoping review on housing and wellbeing. To mark this focus on communities and the built environment, we’re sharing a blog from the English Housing Survey Team in Department for Communities and Local Government.

Earlier in 2016, the Department for Communities and Local Government (DCLG) in collaboration with ZK Analytics and NatCen, published a report on housing and well-being. The report analyses life satisfaction and anxiety as measured by the ONS personal well-being questions. It also uses the English Housing Survey (EHS) to identify the personal characteristics and housing factors associated with wellbeing. Using a series of linear regressions, the report assesses the strength of housing factors in relation to that of personal characteristics.

Because the EHS includes both an interview with the members of the household and a physical inspection of the property, the analysis could examine the impact that housing conditions (e.g. tenure, type, dwelling condition, state of disrepair) as well as personal characteristics have on an individual’s wellbeing. This is the first time that such analysis has been conducted.

Main findings include:

  • Life satisfaction and anxiety are primarily driven by personal characteristics such as health, marital and employment status. For example someone who considered their health to be very good had a level of life satisfaction 2.7 points higher than someone who considered their health to be very bad (on a scale ranging from 0 to 10). Some housing circumstances do however have significant effects.
  • The top housing factor associated with both life satisfaction and anxiety was being in arrears with rent or mortgage payments. Being in arrears reduced an individual’s life satisfaction by 0.6 points and increased anxiety by 0.6 points (on a scale ranging from 0 to 10). These effects were similar in size to some personal factors, such as income after housing costs. This was associated with a difference of 0.6 points in life satisfaction between the lowest and highest income quintiles.
  • Life satisfaction appeared to be influenced by other housing factors, however anxiety did not.
  • For life satisfaction, the second most important property-related predictor was the type of tenure, with social renters having higher levels of life satisfaction (0.2 points higher than outright owners), if all other factors and characteristics were held constant.
  • The next most important property-related predictor of life satisfaction was the type of dwelling. Compared to living in terraced houses, living in flats or semi-detached houses decreased life satisfaction. The largest difference observed was between people living in terraced houses and high rise purpose built flats; their level of life satisfaction was, on average, 0.3 points lower.
  • The final significant property-related predictor of life satisfaction was repair costs. As the average cost of repairs per square metre increased from £0 to £41 (the mean), the level of satisfaction decreased by 0.03 points.
  • Overcrowding (as measured by the bedroom standard) and other housing factors did not appear to have a significant effect on life satisfaction.

The research raises several interesting questions, and has helped us identity a number of areas which would benefit from further exploration.

For example, we are considering adding questions to the survey that assess housing conditions by measuring respondents’ subjective assessments. Respondents could be asked to state, for instance, whether they think their household is overcrowded (as opposed to using an objective indicator such as the Bedroom Standard). In addition, other factors relating to housing and the built environment such as age of dwelling, presence of outdoor space and dwelling density could be explored. The drivers of satisfaction with accommodation and what is the relationship between satisfaction with accommodation and well-being could also be analysed.

More detail, including links to the other reports recently published by DCLG can be found hereEnglish housing survey datasets are also available for analysis. Please contact ehs@communities.gsi.gov.uk for any further queries.

Happy holidays!

winterwellbeingAll of us here at the What Works Centre
for Wellbeing would like to wish you happy holidays with health, happiness and wellbeing for the new year!

We’ve had a great year and you’ve helped us grow in reach to thousands of people. Thank you.

Aside from co-hosting a range of conferences and workshops, in 2016 we launched our first systematic review and measuring wellbeing discussion paper.

We’re looking forward to an exciting 2016: more systematic reviews focussed on the different wellbeing impacts of work, learning, community, sport and culture. We will also be publishing more discussion papers and practical guides for policy makers and practitioners. Follow us to keep updated on our events and publications as dates are announced.

All the best from the What Works Wellbeing team.

Can there be ‘common currency’ to measure wellbeing?

To introduce the Centre’s new Measuring Wellbeing series, our head of evidence and analysis explores the approach set out in our first discussion paper by Prof. Richard Layard on measuring wellbeing and cost effectiveness using subjective wellbeing.

If we want to increase society’s wellbeing, should we build new houses or encourage community choirs? How can we compare the wellbeing benefits of a new crime prevention initiative with an antenatal care programme?

Professor Richard Layard, from the Centre for Economic Performance at the LSE, presents an eloquent, albeit contested, answer. He proposes that we could decide between options by comparing their impact on life satisfaction (on a scale of 0-10), as measured by the Office of National Statistics. Programmes should evaluate their effectiveness, using good quality evaluations, by assessing their impact on life satisfaction. Or, if we already know the impact of a programme on a measure such as the Warwick-Edinburgh Mental Well-being Scale, or job satisfaction, we could convert these to life satisfaction. We may not have precise models to be able to use as an exact exchange rate,  but we can get an indication by by looking at correlation between the measures when they have both been asked in another survey*.

Ultimately, Professor Layard sets out that policies should be converted to life satisfaction. They can then be compared based on their potential increase to life satisfaction for a given amount of money. This would be weighted for inequality so that more importance is given to raising life satisfaction when it is low.

It is a neat solution to a difficult problem. In Government, we make conversions and assumptions all the time: assessing the equivalent monetary benefit of a safe neighbourhood by looking at differences in house prices; asking people how they would compare the risk of death to the risk of breaking their arm; establishing how important road safety is compared to faster speed limits. The results may not be perfect, but they allow us to make a start at comparing options. It is better to be roughly right than completely wrong and not even attempting to make these comparisons.

Professor Layard argues that, rather than trying to convert through monetary values, we should focus directly on what is most important – asking people about their own assessment of their lives.

Overall, how satisfied are you with your life nowadays?

There are, of course, difficulties in choosing life satisfaction as our unit of measurement. Academics have challenged whether people interpret the questions in the same way  and whether we can compare across people. Studies question how responses to satisfaction compare to actual lived experiences. We know that people adapt to their circumstances: individuals are initially happier when conditions improve to give them, for example, less crowded housing and improved physical mobility. But after a while we adapt, and as we become used to these better conditions our expectations increase. Since life satisfaction may be a judgement against the individual’s expectation, it may be difficult to compare even with the same individual over time.

There are also challenges with putting this into practice. While life satisfaction might be associated with all other measures of subjective wellbeing, that does not mean it captures all the variation in other measures. A programme may lead to significant changes in physical or mental health, or satisfaction at work, each of which are important in their own right, but they may not register on a scale of life satisfaction.**

So, what should we do?

Life satisfaction is an important and meaningful indicator of wellbeing. It correlates with other measures of wellbeing, including neurological assessments. However as a Centre, we want to reflect the multi-dimensional nature of wellbeing. We do want to reach a point where we can consistently compare the impacts of different choices with common measures, so we can begin to understand how important certain actions are compared to others. At the same time we want to continue to support, collect and record other outcome measures and understand the value that domain-specific measures and other approaches can bring.***

This discussion series will bring together leading thinkers and practitioners, with their views of how we could define and measure wellbeing and use this in decision-making in different sectors across UK. We want to initiate public discussion and debate about what next – across thinking and practice – to come to a shared, open decision about the methodologies and measurements we need to improve the wellbeing and reduce wellbeing inequalities – of individuals and communities across the UK.

These discussion papers are mainly aimed at analysts, wanting to understand the latest thinking and theoretical underpinnings, however the accompanying blog and ‘Practical Guides’ are aimed at all audiences who may be considering how to put wellbeing into practice.

These scales may be showing something slightly different to life satisfaction, however when both sets of questions are asked in the same survey, it is possible to see how similar they are (how much a change in e.g. WEMWEBS maps onto a change in life satisfaction).  We can assume that this link between the two scales holds in other cases, where only e.g. WEMWEBS has been asked.

** For example, life satisfaction has not been found to be very sensitive to physical health conditions relative to other health-related measures.

*** Some may also support the case for an improvement in existing techniques, so that approaches such as contingent valuation can be more meaningful.


Davidson, R.J. (1992). ‘Emotion and affective style: hemispheric substrates’, Psychological Science, 3(1), pp. 39-43.

Dolan, P., Kudrna, L.,  and Stone, A. (2016). ‘The measure matters: an investigation of evaluative and experience-based measures of wellbeing in time use dataSocial Indicators Research. ISSN 0303-8300

Haybron, D.M. (2007). ‘Well-being and Virtue’, Journal of Ethics & Social Philosophy, 2 (2)

Mukuria, C., Peasgood, T., Rowen, D. and Brazier, J. (2016). ‘An empirical comparison of well-being measures used in UK’. Research Report RR0048: University of Sheffield and the University of York.

Ponocny, I., Weismayer, C., Stross, B. et al.  (2016) ‘Are Most People Happy? Exploring the Meaning of Subjective Well-Being Ratings’ Journal of Happiness Studies 17: 2635 doi:10.1007/s10902-015-9710-0

Ralph, K., Palmer, K. Olney, J. (2011). ‘Subjective Well-being: a qualitative investigation of subjective well-being questions’ (Working paper for the Technical Advisory Group on 29 March 2012)