Wealth Without Wellbeing

When growth stopped buying life — and how an economy tips, over time, from integration to extraction.

The foundational hypothesis of this programme is simple: capability outran integration. A society can keep getting richer, cleverer and more productive while the thing that holds it together — trust, shared direction, the health of its bonds — stops keeping pace, and then begins to drain. The claim has two halves:

• First, it caps. Beyond a point, more capability stops buying more life. Wealth keeps rising; wellbeing flattens.
• Then it hollows — out of sight. Not because wealth runs out of use, but because the measure of it — GDP, growth, the balance sheet — stops being a means to wellbeing and becomes the end. Once the proxy is the goal, chasing it starts to consume the very wellbeing it once tracked — while the society keeps its confident self-image and the substance drains underneath.

This essay tests both halves against the record, across several countries — and then asks the harder question: if an economy can tip from integration to extraction, when did ours?

It is worth saying plainly at the start what many people feel privately: that the economy has been extractive for a long time — and yet that within a single lifetime it was noticeably different. Both are true. What changed was not a switch thrown on one date but an acceleration of a long-present tendency: part deregulation and financialisation, part a drift in what institutions were oriented toward, and part a shift in what a whole culture came to prize — status and accumulation over contribution and belonging.

The cap — across countries, not just one

Start with the honest version, because the easy one is wrong. More money does, on the whole, track more wellbeing: richer and more equal countries report higher life satisfaction, and within any country the better-off are happier than the poor. The interesting truth is about shape. Wellbeing rises with the logarithm of income — each further doubling buys less than the last — and across whole nations over time the gains largely flatten. That is not a single-country quirk:

Two honest caveats the table cannot hold: Japan's apparent flatness has been challenged as partly an artefact of changing survey wording (Stevenson & Wolfers), and the Easterlin paradox is contested in general. What appears to survive most versions of the dispute is the steep diminishing returns, not a literal flat line.

The Nordic row is the one that matters most — and it must be read honestly, because it cuts both ways. If wealth itself hollowed societies, the richest egalitarian states would be hollowing fastest; instead they have sat at the top of the wellbeing rankings for years. Through this lens the difference reads not as how much they have but as how it is held — decades of deliberate low inequality and intact social bonds, a settlement that, in this programme's terms, reads as a relatively integrative one. That is the thesis in one line: it is not wealth that drains a society, it is the extractive orientation. But the Nordics are not a clean win, and to present them as serenely stable would be the cherry-pick. Their young, too, have begun to slip — part of the same youth-wellbeing decline now seen across the wealthy world, with the attention economy implicated — because no national settlement, however integrative, can fully wall itself off from a global economy tilting toward extraction. They held longer and held higher; they have not escaped. That is what the lens would predict: orientation is protective, but it is not a force field.

The turn — when wealth starts costing

What turns a plateau into a decline is not the wealth; it is what a society makes of the measure of it. A good servant — GDP, growth, the share price — becomes a bad master the moment it stops being a gauge of wellbeing and becomes the thing pursued for its own sake, the bond a society fastens its identity and ideology to. Then proxy and purpose come apart, and optimising the proxy starts to cost the purpose. That is why the slide shows first and hardest not where wealth is scarce, but where the measure has most completely displaced the purpose it was meant to serve. It is not a question of which society is richest — Western Europe is about as rich as the United States and, on the whole, has not slid as far — but of which has let the number crowd out the thing it was supposed to track. The society furthest down that road — measured independently, by inequality and the financialised share of the economy, not by the wellbeing outcome we are trying to explain — is the most market-absolutised, not the wealthiest; a comparably rich society that kept the measure in its place is the case the reading has to pass, and on present evidence does.

In several of the richest societies the curve has lately gone past flat, to falling. The United States has slid to its lowest-ever place in the world happiness rankings — from 15th in 2023 to 23rd in 2024 to 24th in 2025 — its lowest yet — and the fall is concentrated among the young; the same downturn appears across the wealthy English-speaking world and parts of Western Europe, but not in the Nordics. Alongside it: deaths from suicide, alcohol and drugs in the US more than doubled in two decades (from roughly 65,000 a year in the mid-1990s to over 150,000 by 2018), and loneliness is now treated as a public-health emergency. The causes are genuinely contested — material insecurity for some, social media, the thinning of community and attention — and honesty requires holding that openly. But the shape is what the lens predicts: capability still climbing while the things that make life worth living turn down.

When did the economy tip? — the two clocks

The programme's way to date such a turn is the two clocks. One is the capability clock — output, productivity, markets, the things that can be counted. The other is the integration clock — median wages, the health of bonds, wellbeing. While a society is cohering, the two run together. The tip into extraction is the moment they diverge: the capability clock races ahead while the integration clock stalls. Note what this is and is not. It is not a society announcing that it now serves extraction — comfortable decline never looks like that. It is masked: the self-image and the stated purpose stay intact while the substance hollows, visible wealth and output running ahead of a thinning inner capacity. That gap — the outward showing running ahead of the inner substance — is the framework's signature of comfortable decline, and in the United States the divergence is unusually well documented.

Read this way, the estimate is clear and it matches the lived memory: the orientation began flipping extractive at the 1970s hinge, accelerated through the 1980s, and the damage to wellbeing only became visible around the 2010s. The lag is not a puzzle; it is exactly what the framework expects. A rich society runs on deep reserves — institutional, financial, social — and those reserves give a long runway on which a hollowing system keeps looking healthy. "Nominal high, effective collapsed" is the framework's phrase for it, and it is why the most "sudden" failures are the ones that were draining quietly for decades.

It also explains why no single cause settles the argument, and why your own memory is right that it was "different" once. The turn was overdetermined: deregulation and financialisation changed the rules; the orientation of finance, firms and even governments drifted from provisioning toward extraction; and the surrounding culture came to measure a life by status and accumulation rather than contribution and belonging. Three currents, one direction. The postwar decades felt different because, for a few decades, the currents ran the other way.

A note on the figures: each is debated at the edges — the productivity–pay gap partly on which inflation measure is used and whether non-wage benefits are counted; the happiness series on survey method. The direction and the timing, though, are robust across measures.

What the model adds — and what it does not yet

The contribution of the lens is not the dates — economists found those. It is the reading — the programme's central hypothesis, not a settled finding — that the productivity–pay gap, the rising top share, the flat-then-falling wellbeing, and the deaths of despair are not four separate problems but one structural event seen from four windows — capability decoupling from integration, with the cost exported onto wages, the young, and the social fabric. And this is the part to be clear-eyed about: it is the same structural signature the programme reads across very different systems — a failing institution, a hollowing dynasty, a civilisation past its peak — each read put to the same kind of sealed test — and it is that recurrence, one pattern surfacing domain after domain, that the reading rests on, not this single case. The economy is one application of it, here at the scale of a whole economy.

Where this stands in the engine, honestly. The two-clocks reading here is a structural reading of the historical record, not yet an engine result: we ran it more than once, across successive rebuilds of the engine, and it kept coming back negative on the hard question. Our reading is not that the case is wrong but that the first attempts looked for the wrong thing — for a society's orientation to openly flip toward extraction, when in life a comfortable society never announces that. It keeps believing it is doing fine while the substance hollows underneath — the outward showing running ahead of the inner substance, the signature the framework calls masked decline. The engine already reads a close cousin of it elsewhere (the 2008 ratings failure, where the form was kept while the function hollowed). What is pending is narrow: the economy-specific run that would pin the pattern to this case with a sealed result. A negative there would not overturn the wider framework — which rests on several independent applications, each under its own sealed test — but it would weaken one line of evidence; and whether the recurrence seen across institutions, dynasties and eras is one underlying structure or several that merely resemble each other stays an open question the programme tests case by case, not one it assumes.

The trail in the world — the charter gap, and a first blind test

The engine's own behaviour pointed us here. Driven hard, the model would not let a healthy society's stated orientation flip — and that is not a flaw, it is the world: organisations' declared vision, mission and purpose do not decay across decades — if anything they grow loftier, fuller of stakeholder and public-good language; the words stay noble. So the decline cannot live in the stated compass. It lives in the gap between what a system says it is for and what it actually does — widening unseen while the self-image holds. That gap is the framework's candidate signature of masked decline, and it is the quantity the engine is built to locate: not a vague malaise but a datable opening between profession and conduct, readable system by system on the canvas.

Which makes the case testable in the world, on three readings that should move together: stated purpose — does the declared compass hold? conduct — is extraction rising (pay decoupling from output, gains concentrating)? and outcome — is wellbeing falling even as GDP climbs? The prediction is specific and unforgiving: the declared compass holds high or climbs, extraction rises, wellbeing falls.

One leg has now met a blind, sealed test — registered before scoring, the kind that earns trust. We asked whether the decline shows up in the words: whether company purpose-language has homogenised across four cohorts (about 300 firms each) spanning 1996 to 2024. The registered prediction — a steady narrowing toward a common script — was only partly borne out. The language did narrow a little, and the narrowing held up against size and disclosure-length controls; but not steadily — it widened again in 2014 — so it failed the pre-registered monotone test, and with only four cohorts the power is weak. Suggestive at most. The honest call was to retire that word-based measure — the orientation inside the language carried no clean trend — while noting that the declared language itself, if anything, grew loftier across the period. So the say-do gap does not live in whether the words are noble; it lives in the distance between those rising professions and what conduct actually does — which is where the framework said to look. Set beside the conduct leg (the long productivity–pay decoupling and the concentration of gains) and the outcome leg (wellbeing flat-then-falling against rising output), the three readings are, so far, mutually consistent — but consistency is not yet support for a single mechanism. What stays a structural reading, not yet validated, is the claim that ties them together: that the widening gap is the decline. This is live research, not a finished result — and, in the end, it is the question worth answering: where a society went off the rails, and where it is heading.

How this could be wrong

State the falsifiers plainly. If the recent wellbeing decline turns out to track pure material deprivation everywhere — falling only where incomes fell — then it is an income story, not an orientation story, and the lens adds nothing. If the low-inequality, more-integrative rich societies (the Nordics) hollow at the same rate as the extractive ones once the data is in, the distinction the whole argument rests on collapses. And the integrated claim — that the widening gap between professed purpose and conduct is the decline — has a sharp, checkable form: the gap must predict the outcome. Across cases, the systems with the widest and fastest-widening say-do gap should fare measurably worse on wellbeing than those without; if a wide gap carries no such penalty, the gap is a coincidence, not a mechanism. That is the forward test, and it can fail. Each of these is checkable; each was stated before the fact.

An ongoing research project, built on public data and read through the programme's two-clocks lens: the historical figures are cited; its most testable leg now has data; the claim that ties them into one mechanism is a structural reading under forward test — not yet a validated result.