What the data actually shows
The relationship is real but logarithmic. The most reliable finding across studies is that happiness rises with income, but in proportion to the logarithm of income — meaning it takes ever-larger raises to produce the same bump in wellbeing. Going from $30,000 to $60,000 tends to do far more for a person than going from $300,000 to $600,000, even though both are a doubling.
There is a genuine, unresolved debate about whether the effect plateaus. The widely cited Kahneman and Deaton study (2010, PNAS) found that day-to-day emotional wellbeing rose with income only up to roughly $75,000 a year and then flattened, while broader life evaluation kept climbing. Killingsworth (2021, PNAS), using real-time experience sampling, found no such plateau — experienced wellbeing kept rising with log income well past that point. Rather than one side winning, the two camps ran a rare adversarial collaboration (Killingsworth, Kahneman and Mellers, 2023, PNAS) and concluded that for most people happiness does keep rising with income, but a flattening appears mainly among an unhappy minority. So the 'magic number' framing is a simplification of a more careful picture.
Zoom out, and the picture gets weaker still. The Easterlin paradox observes that although richer people within a country tend to be happier than poorer ones, decades of rising national income have not reliably raised a country's average happiness. Part of the reason appears to be relative: research by Luttmer (2005, QJE) found that people report lower wellbeing when their neighbours earn more, holding their own income constant. Much of what money 'buys' in happiness terms seems to be position relative to others, which is a moving target.
Real numbers in context
The famous threshold was specific and narrow. Kahneman and Deaton (2010) found day-to-day emotional wellbeing flattened around $75,000 a year — a single U.S. measure from over a decade ago, for emotional wellbeing only; life evaluation kept rising with income. Killingsworth (2021) found no plateau at all using real-time sampling, and the 2023 adversarial collaboration concluded happiness generally keeps rising with log income, with flattening mainly among the least happy. Treat any single 'happiness number' with caution.
The relationship is logarithmic, which is the part most worth internalising: because wellbeing tracks the log of income, a $10,000 raise means far more to a household earning $40,000 than to one earning $400,000. And the Easterlin paradox is a reminder that at the national level, decades of rising income have not reliably raised average happiness — so the within-country, relative, diminishing-returns picture is the one the data best supports.
The 'money and happiness' headlines usually collapse several distinct findings into one slogan. These are the load-bearing results, each measuring something slightly different.
| Study / finding | What it measured | Result |
|---|---|---|
| Kahneman & Deaton (2010) | Day-to-day emotional wellbeing | Flattened around ~$75,000/yr; life evaluation kept rising |
| Killingsworth (2021) | Experienced wellbeing, real-time sampling | No plateau — kept rising with log income |
| Killingsworth, Kahneman & Mellers (2023) | Adversarial re-analysis | Happiness generally keeps rising; flattening mainly among an unhappy minority |
| Easterlin paradox | National income over decades vs average happiness | Weak — rising income hasn't reliably raised average happiness |
| Luttmer (2005) | Wellbeing vs neighbours' income | Lower wellbeing when neighbours earn more, own income held constant |
Why this feels different from how it actually is
Money feels like it should matter more than the data says because we are very good at imagining the pleasure of a purchase and very bad at imagining how quickly we will get used to it. We picture the new house or salary and forget that our expectations rise to meet whatever we get — a pattern so reliable researchers gave it a name, the hedonic treadmill.
It also feels different because the comparison never stops. A raise that would have thrilled you five years ago feels ordinary once it becomes your baseline and once you can see what the next tier up has. Since relative standing drives a real share of the effect, getting richer often moves the reference point along with you, leaving the felt gap roughly where it was.
And the cultural messaging is relentlessly one-directional. The visible, marketed version of the good life is almost entirely about acquisition, so the modest, slow, diminishing-returns reality of money's actual contribution to wellbeing is easy to lose. The result is a persistent sense that the next financial milestone is the one that will finally settle things — a sense the evidence does not support.
Money helps, it helps less than people expect, and beyond covering your needs and a buffer it is one factor among several rather than the decisive one.
What the research says to do about it
Where money clearly helps wellbeing is in removing strain rather than in accumulating beyond your needs. The strongest part of the income-happiness curve is at the lower end, where additional income relieves the day-to-day stress of insecurity. Building enough of a buffer to stop worrying about ordinary shocks tends to do more for wellbeing than the same money applied higher up the curve.
How you spend appears to matter more than how much. A line of research suggests buying experiences rather than possessions yields more durable satisfaction (Van Boven and Gilovich, 2003), partly because experiences resist comparison and become part of your identity and memories rather than fading objects. The effects are modest, but they point in a consistent direction.
Spending on other people also shows a small but reasonably consistent positive effect. Dunn, Aknin and Norton (2008, Science) found that people directed to spend money on others reported higher happiness than those who spent it on themselves. Buying back time — paying to offload tasks you dislike — is another candidate with supportive evidence. None of these are large effects, but they are cheaper and more reliable than chasing the next income tier.
What the research says does not help
Treating a specific income figure as a finish line does not help. The plateau debate is unresolved, the numbers are dated and country-specific, and the famous '$75,000' was never a universal threshold — it was a 2010 U.S. figure for one measure of wellbeing. Building your plan around a magic number sets you up to feel that hitting it should have fixed something it was never going to fix.
Buying status to win the comparison reliably underdelivers, because relative standing is a treadmill. Since wellbeing partly depends on income relative to your reference group, climbing pulls your reference group up with you, so the felt advantage erodes. Spending aimed at out-positioning others tends to buy the briefest satisfaction of all.
Assuming more money will do nothing is also wrong, and the cynical version of this research is as misleading as the materialist one. For people under financial strain, more income genuinely and measurably improves life. The data does not say money doesn't matter; it says it matters with diminishing returns and is easy to overweight, which is a very different claim.
The '$75,000' was never a universal threshold — it was a 2010 U.S. figure for one measure of wellbeing.
What this looks like in real life
A $10,000 raise, two very different households
Because wellbeing tracks the logarithm of income, the same $10,000 raise does far more for a household earning $40,000 than for one earning $400,000. For the first, it can lift real day-to-day strain; for the second, it barely registers against the baseline. The dollar figure is identical; the wellbeing return is not — which is the whole point of 'diminishing returns.'
Hitting the number that was supposed to fix things
Someone spends years treating a target income as the finish line, reaches it, and feels the same. That isn't a personal failure — it reflects both the hedonic treadmill (expectations rise to meet whatever you get) and the fact that relative standing moves with you. The research argues against building a plan around any single 'magic number,' because it was never going to settle what people hoped it would.