Nick Gellatly
How economists think about happiness
Boots
On of the most useful things in this week’s session was not from an economist. It came from Nic Gibson in the chat, and it was Terry Pratchett.

In Men at Arms, Sam Vimes works out why the rich are so rich. A good pair of leather boots costs fifty dollars and lasts ten years. Cheap boots cost ten dollars and are leaking by the autumn. So the man who can afford fifty dollars has dry feet for a decade, and the man who cannot has spent a hundred dollars over the same period and still has wet socks.
It is a joke, and it is also the entire week in miniature. Poverty is expensive. Money buys more than things, it buys the ability to make the economical choice. Any account of happiness that treats income as simply a number has missed the mechanism.
This week asked how economists think about happiness, which turns out to be a question about what we count, who does the counting, and what happens when the counting stops.
What the economists did

Dr Imko Meyenburg’s lecture traced a shift in method. Traditional economics inferred welfare from choices. It watched what people bought and how they worked, assumed they were maximising utility, and read happiness from what people were able to buy. The economics of happiness does something simpler and stranger: it asks people.
Large surveys now collect self-reported life satisfaction and treat it as a proxy for welfare. It is a broader measure than consumption and a softer one, and the whole field lives in that trade-off.
What the evidence says about money. Income matters, and it matters most where it relieves poverty and meets basic needs. Beyond that, the returns diminish, for two reasons this module has met before. Hedonic adaptation means we get used to improvements. Social comparison means satisfaction depends on where we stand relative to other people rather than on what we have in absolute terms.
What else matters. Health, stable employment, secure relationships, freedom, trust and economic stability. Unemployment, inflation, divorce and insecurity all reduce wellbeing. None of these is a mood, and most of them are conditions rather than choices.
The Easterlin paradox (also mentioned in the comments) sits at the centre of the field. Within a country at any given moment, richer people report greater life satisfaction. Over the long run, as a country’s average income rises substantially, its average happiness stays broadly flat. Easterlin observed this in the United States in 1974 and it has been argued about ever since.
The leading explanation is social comparison. If everyone’s income rises together, nobody gains relative standing, so the collective gain is smaller than the individual one appeared to promise. Adaptation is the second explanation: the lift from extra money fades as the new standard becomes normal.
It is worth being honest that this is contested rather than settled. Newer research does find that richer countries tend to be happier on average, and some economists think the long-run relationship is weaker or more complicated than Easterlin first claimed. What survives the argument is narrower and still important: growth does not automatically convert into wellbeing, and how growth is shared may matter as much as how much of it there is.
I am glad I am not studying economics. It is more complicated than I expected, and rather more interesting.
So how happy are we?

The seminar asked how happy people are in your country, which sent me to the data afterwards.
The Office for National Statistics has asked the same four questions since 2011, each scored from nought to ten. How satisfied are you with your life nowadays. To what extent do you feel the things you do are worthwhile. How happy did you feel yesterday. How anxious did you feel yesterday. The first three run the way you expect. Anxiety is scored so that lower is better, which catches people out.
The last full annual release, for the year ending March 2023, reported life satisfaction at 7.45, worthwhile at 7.73, happiness yesterday at 7.39 and anxiety yesterday at 3.23. All four had moved in the wrong direction on the previous year and all four were below pre-pandemic levels. Nearly one adult in four, 23.4 per cent, reported high anxiety.
What drives it is the part that matters for this module. The ONS regression puts self-reported health a long way ahead of everything else, then marital status, then employment and education. Disabled adults reported poor wellbeing at roughly five times the rate of non-disabled adults. The ONS is also admirably honest that its models explain only between 11 and 22 per cent of the variation between people, so most of what moves a life is not in the data at all.
And then the thing I did not expect.
That release is from November 2023. There has not been another one.
In October 2024 statistics from the Annual Population Survey lost their accredited status because of problems with response rates, and the ONS paused the Personal Well-being in the UK publication. Quarterly headline estimates continue, but the detailed annual picture, with all the breakdowns by characteristic, has not appeared for nearly three years. The What Works Centre for Wellbeing, which existed to turn this evidence into policy, closed in April 2024 because nobody would fund it.
So the honest answer to how happy people are in the UK is that we used to know, roughly, and we have rather stopped checking. For a module about what happiness means, that is not a footnote. It is the story.
Which country, though

The seminar discussion turned on country of origin, which made me set out a position I usually leave vague. I am an Anglo-Scot. Born in Scotland to a Scottish father and an English mother, educated in Edinburgh, living in Cambridgeshire, with the family home still in Edinburgh. When someone asks how happy people are in my country, so I looked at both.
The scores are almost the same. Scotland tracks the UK closely on the ONS measures, and the differences between the four nations are small and mostly not statistically significant. In the last full year wellbeing fell in England, Wales and Northern Ireland and rose slightly in Scotland, but only England’s change was statistically significant, and the devolved nations have smaller samples and wider uncertainty. Anyone telling you Scotland is measurably happier than England is overreading the evidence, and I say that as someone who would quite like it to be true.
The frameworks are not the same at all. Scotland has had a National Performance Framework since 2007, aligned to the UN Sustainable Development Goals, with national outcomes rather than growth targets as the stated purpose of government. It is a founding member of the Wellbeing Economy Governments partnership alongside Iceland, New Zealand, Finland and Wales, and it publishes a Wellbeing Economy Monitor. On paper it is one of the more serious attempts anywhere to govern by something other than GDP.
And then the gap. The promised Wellbeing and Sustainable Development Bill was in the Scottish Government’s 2021-22 programme, consulted on in 2024, and not brought forward. It eventually reached Parliament in March 2025 as a member’s bill from Sarah Boyack MSP, creating a duty on public bodies to have due regard for wellbeing and sustainable development, and establishing a Future Generations Commissioner. WEAll Scotland’s own assessment is that progress has stalled and that Scotland risks falling behind an agenda it helped start.
The comparison is useful because it flatters nobody. Scotland has the better framework and roughly the same outcomes. Either frameworks do not do very much, or this one was never given teeth. I suspect the second, and the difference between due regard and a duty to deliver is exactly where it was lost.
Wales, meanwhile, has had the statutory duty and the Commissioner since 2015, and is the only part of these islands where the question has an answer with legal force behind it.
The people who are not in the data
The third comparison I wanted was LGBTQ+ wellbeing against the general population. What I found was mostly an absence.
The ONS first broke wellbeing down by sexual identity in 2017, using three years of data. Lesbian, gay and bisexual people rated themselves lower than heterosexual people on life satisfaction, worthwhile and happiness, and higher on anxiety. Bisexual people had the widest gap, with around 30 per cent reporting high anxiety against 19.5 per cent of heterosexual people. People who answered “other”, or declined to answer, also reported lower wellbeing than average. A later meta-analysis of twelve UK population health surveys found the same pattern, and research here at ARU found older LGB people reporting lower quality of life than their heterosexual peers even after adjusting for other factors.
That 2017 article is nine years old. There is no equivalent series that has been kept up. The 2018 National LGBT Survey, with 108,000 responses, has not been repeated. Trans people appear in almost no official wellbeing statistics at all, because sample sizes in general population surveys are too small, which means the evidence base is charity research rather than official data. And the ONS release that would carry these breakdowns is the one currently paused.
The context has not stood still while the measuring did. Police recorded 19,127 sexual orientation hate crimes and 3,809 transgender identity hate crimes in England and Wales in the year ending March 2025, down 2 and 11 per cent on the year before. Those falls are harder to celebrate than they look. The Government’s own National LGBT Survey found that fewer than one in ten LGBTQ+ people report hate crimes, and Stonewall argues that falling reports partly reflect falling trust in the police rather than falling hatred. The figures also predate the April 2025 Supreme Court judgment, and Stonewall reports a rise in fear in the period after it. In February 2026 the Government tabled an amendment to the Crime and Policing Bill to make anti-LGBTQ+ hate crime an aggravated offence.
So: we know LGBTQ+ people report lower wellbeing. We do not know by how much today, because nobody is checking. A group whose wellbeing gap is attributed to discrimination, and whose data collection then quietly lapses, is a worked example of the question this module keeps returning to. Whose happiness counts? Apparently the happiness of people we are still counting.
Who decides what counts as progress
The group activity asked who decides what counts as progress, and whether happiness is political. Four weeks ago I would have said the first was a technical question. It is not.
Statisticians decide what is measured. The four ONS questions were a choice. They capture how life feels from the inside, which means they cannot see what Sen and Nussbaum call capabilities, what a person is actually able to do and be. They also cannot see adaptation. Somebody who has lowered their expectations may report contentment with a life that is objectively constrained, which is Felicity from Week 2 turning up in a dataset.
Statisticians also decide what stops being measured, and that is the quieter power. The LGB breakdown has not been refreshed since 2017. The annual release is paused. The What Works Centre is gone. Each decision was defensible on its own terms. Together they are a choice about which inequalities stay visible.
Governments decide what the numbers are for. New Labour said in 1998 that wellbeing would steer policy. David Cameron launched the ONS programme in 2010. But a measure with no duty attached is a dashboard, not a lever. Wales has the duty. Scotland has the framework. Westminster has neither.
International bodies decide the league table. The World Happiness Report, published under UN auspices since 2012, ranks countries on a single life-evaluation question. It made wellbeing globally legible and simultaneously made it competitive, which is a strange thing to do to an emotion. As Seymour pointed out in last week’s reading, an idea inspired by Buddhist Bhutan arrived wearing the measurement culture it was meant to replace.
Media and platforms decide what registers as a problem. A ranking is a story. A pause in a statistical series is not. Which is why you have heard about one and not the other.
And the people being measured decide almost nothing. Carnegie UK made exactly this criticism of Scotland’s framework: people have not been asked what wellbeing means to them, or what they think should be counted.
Is happiness political? Unarguably, and the history settles it. Seymour’s chapter traces happiness from providential gift to political project, written into the American Declaration and the French Revolution, then claimed in turn by liberalism, communism and fascism, each with its own account of how human beings ought to flourish. Under Stalin, failing to notice the happiness conferred by Soviet citizenship became effectively a political crime. Mussolini and the Nazis both organised leisure in the service of the state. The idea that happiness is a private matter is itself a political position, and a recent one.
The current framings line up roughly like this. Neoliberalism treats happiness as individual and best served by growth and free choice, which critics such as Davies, Cabanas and Illouz argue converts structural problems into personal failings. Environmentalism treats it as capped by planetary limits, which raises the question of who bears the cost of staying inside them. Nationalism treats it as belonging and continuity, which takes social connection more seriously than most and is also the framing most prone to deciding that some people’s belonging requires other people’s exclusion. Populism treats it as something stolen by an elite, which names real losses of control and status and rarely addresses their causes.
Edoardo. In the breakout room I talked with Edoardo, an Italian student here at ARU. He said he feels welcome in this country and is frightened of Reform.
So am I.
That is the conversation I keep thinking about, because it will not appear in any dataset. Two people, from different countries and possibly different generations, describing the same concern. The ONS can tell you that anxiety is running at 23.4 per cent. It cannot tell you that a cheerful Italian student and a late-career Anglo-Scot sat in a Teams breakout room and recognised each other’s fear immediately.
I want to be careful about the question of people supporting policies against their own interests, because the phrase assumes I know someone’s interests better than they do. The research offers a few answers: status threat predicts support for radical-right politics better than income does, people weigh loyalty and belonging alongside material self-interest, and grievance is vivid in a way that structural explanation never is. All true. It is also true that people may simply value sovereignty or cultural continuity above GDP, and calling that false consciousness is both condescending and a way of avoiding the argument. What I will say plainly is the narrower point: LGBTQ+ wellbeing is unusually sensitive to political weather. The Home Office itself has suggested that increased public discussion of trans issues contributed to recorded hate crime. For some of us, happiness is not political in the abstract. It is political most weeks.
Growth, or something else
The lecture’s policy implication is that GDP growth alone is an incomplete measure of national success. Bhutan is the standing example of the alternative.
Gross National Happiness rests on four pillars, sustainable and equitable development, cultural preservation, environmental protection and good governance, measured across nine domains including psychological wellbeing, health, education, time use, community vitality, living standards and ecological resilience. The important move is that it treats happiness as a public policy outcome rather than a private mood. An economy should serve health, community, nature, fairness and freedom as well as income.
It also comes with a warning that I think matters more than the model. Bhutan’s GNH has been criticised for allowing cultural preservation to justify exclusionary policy. So the transferable part is the method, broad multidimensional measurement, and not the assumption that a national happiness agenda is automatically a progressive one. A country can measure its collective wellbeing carefully and still decide that some people are outside the collective.
Which is the strongest argument I have found this term for measuring minorities separately. A national average can disappear a minority entirely. If Scotland combined its existing framework with explicit equality monitoring, that would be the serious version.
Should nations prioritise happiness over economic growth?
The case for is the Easterlin paradox, the fact that GDP counts a traffic jam and a hospital stay as positives, and the existence of working alternatives in Bhutan, Wales and New Zealand that have not collapsed. Ruskin got there first and more briefly: there is no wealth but life.
The case against deserves better than a straw man. Growth pays for the health service, and health is the single strongest predictor in the ONS data, so a country that stops growing may find its wellbeing policy unaffordable. The 2023 adversarial collaboration between Killingsworth, Kahneman and Mellers largely overturned the idea of an income plateau, which weakens the strongest version of the Easterlin case. Wellbeing measures are softer than GDP and easier to game, and a government graded on how people say they feel has an incentive to manage the mood rather than the conditions. There is even a democratic objection: GDP does not pretend to tell you what a good life is, whereas a wellbeing framework necessarily does.
My answer is that the dichotomy is the wrong shape. The question is not growth or happiness. It is what growth is for, and whether anything other than growth carries a legal duty. On present evidence the UK has neither a wellbeing duty nor functioning wellbeing statistics, which means it is not currently asking the question at all.
What this does to the café
Queerly Beloved Café has a date. Monday 30 November, and the room booking form has gone in. I’m hoping it will be confirmed soon.
The date carries three things at once, which is more than I planned and better than I deserved. It is St Andrew’s Day, which sits rather well with a week that made me declare myself an Anglo-Scot in a Teams breakout room. It is the first Monday of Advent, the season of waiting in the dark for something that has not arrived yet, and the one stretch of the church year that does not ask anybody to be cheerful. And it is the eve of World AIDS Day.
That last one changes the quiet corner. I had been defending it as a hedge against toxic positivity, somewhere for people finding December hard. On the night before World AIDS Day, in a room built for queer joy, a corner with a chair and a candle is not a caveat. It is the point. A community that has buried a generation does not need to be told that joy and grief belong in the same room. The first of December itself was not available, because I am part of a live brief with the Immersive Storytelling module that day, and I find I prefer the eve anyway. Vigils happen the night before.
The data changes something else. I have been arguing since Week 1 that belonging is a structural answer to unhappiness rather than a personal resilience one. The ONS backs that up from an unexpected direction, since health, relationships and security do the heavy lifting while individual mindset barely features. But the stronger claim is now available. The state has stopped counting LGBTQ+ wellbeing. The last proper look was 2017. So when the noticeboard asks people to send back a photograph, a poem, a complaint or a line on a card, that is a small, unscientific, hopelessly self-selecting act of counting. All those caveats stand. It is still more than is currently happening nationally.
Carnegie UK’s criticism of Scotland’s framework was that people had not been asked what wellbeing means to them. In this room, they will be. I need to write a questionnaire for attendees before the evening, covering what happiness means to them and whether the night did what it set out to do, to sit alongside the photographs, the film and the audio.
And the GNH caveat keeps me honest about my own project. A happiness agenda is not automatically a generous one, and a room that defines belonging can define people out of it just as easily as in. The welcome has to be unconditional in belief and clear about conduct, which remains the hardest design problem I have.
One line from all this that I think is the truest thing I have written this term: the state has stopped counting us, so we will have to count ourselves, and we will do it with cake.
References
Carnegie UK (2024) Exploring Scotland’s National Outcomes. Dunfermline: Carnegie UK. Available at: https://carnegieuk.org (Accessed: 14 October 2026).
Easterlin, R.A. (1974) ‘Does economic growth improve the human lot? Some empirical evidence’, in David, P.A. and Reder, M.W. (eds.) Nations and households in economic growth. New York: Academic Press, pp. 89-125.
Home Office (2025) Hate crime, England and Wales, year ending March 2025. London: Home Office. Available at: https://www.gov.uk/government/statistics/hate-crime-england-and-wales-year-ending-march-2025 (Accessed: 14 October 2026).
Killingsworth, M.A., Kahneman, D. and Mellers, B. (2023) ‘Income and emotional well-being: a conflict resolved’, PNAS, 120(10). doi: 10.1073/pnas.2208661120.
Meyenburg, I. (2026) ‘How economists think about happiness’. Lecture delivered as part of the Ruskin Module MOD010539, Anglia Ruskin University, 14 October.
Office for National Statistics (2017) Lesbian, gay and bisexual people say they experience a lower quality of life. Available at: https://www.ons.gov.uk/peoplepopulationandcommunity/culturalidentity/sexuality/articles/lesbiangayandbisexualpeoplesaytheyexperiencealowerqualityoflife/2017-07-05 (Accessed: 14 October 2026).
Office for National Statistics (2023) Personal well-being in the UK: April 2022 to March 2023. Available at: https://www.ons.gov.uk/peoplepopulationandcommunity/wellbeing/bulletins/measuringnationalwellbeing/april2022tomarch2023 (Accessed: 14 October 2026).
Pratchett, T. (1993) Men at Arms. London: Victor Gollancz.
Royal Government of Bhutan (no date) Gross National Happiness. Available at: https://www.mfa.gov.bt/rbedelhi/bhutan-at-glance/gross-national-happiness/ (Accessed: 14 October 2026).
Ruskin, J. (1862) Unto this last: four essays on the first principles of political economy. London: Smith, Elder and Co.
Scottish Parliament (2025) Wellbeing and Sustainable Development (Scotland) Bill. SPICe briefing SB 25-48. Available at: https://www.parliament.scot/chamber-and-committees/research-prepared-for-parliament/research-briefings/2025/10/8/sb-2548 (Accessed: 14 October 2026).
Seymour, M. (2023) ‘Global happiness: from providential moments to hedonic treadmills?’, in Barclay, K. and Stearns, P.N. (eds.) The Routledge history of emotions in the modern world. Abingdon: Routledge, pp. 29-43.
Stonewall (2026) Hate crime. Available at: https://www.stonewall.org.uk/shaping-policy/hate-crime (Accessed: 14 October 2026).
Wellbeing Research Centre, University of Oxford, Gallup and UN SDSN (2026) World happiness report 2026. Available at: https://data.worldhappiness.report/table (Accessed: 14 October 2026).
Illustrations for this series are made in Canva, in a flat vector style borrowed from broadsheet long-read journalism. There are no people in any of them, which is deliberate: a module about who gets pictured as happy is not well served by stock photographs of contented strangers. AI tools were used for research, drafting support and image generation. The words, the arguments and the responsibility are mine.

















































