Why Money and Work Matter: How People Earn, Why Success Stories Mislead, and What Research Says About Happiness

Start hereIn 2010 Kahneman and Deaton found that day-to-day emotional well-being showed no further progress beyond about $75,000 a year. In 2021 Killingsworth found no such plateau. Can both be right?

This page starts with the words money and economy, then looks at how people around the world earn a living and why success stories can mislead. It ends with what research says about money and happiness, and a few ways to read money claims. It describes findings; it does not tell you what to do with your money.

011. A simple example

Imagine that ten friends each open a café in the same year. Five years later, a magazine interviews the owners who are still open about their secrets. Nobody interviews the ones who closed.

Whatever the survivors did now looks like a recipe, even if the closed cafés did the same things. Keep that picture in mind for section 4.

022. What the words mean

The word economy comes, through French and Latin, from Greek oikonomia. In the standard dictionary of ancient Greek, its first meaning is management of a household or family, husbandry and thriftkonomi/a). Running a state appears later in the entry, as administrationkonomi/a).

The English word money came through Anglo-French from Latin moneta, mint or money. In a standard Latin dictionary, Moneta was a surname of Juno, in whose temple at Rome money was coined. By extension the word came to mean the mint, and coined money.

Salary has a salty story. A standard Latin dictionary says salarium originally meant the money given to soldiers for salt. Merriam-Webster says that idea has no basis in ancient sources: it is an inference from the word’s form, found in that 1879 dictionary and older ones. Pliny the Elder wrote that salt played its part in the honours given for success in war, and that the word salarium comes from it.

Can't confirm Roman soldiers were paid in salt. The sources we read link the word to salt; none says soldiers were paid in salt, and Merriam-Webster says even the salt-money allowance has no ancient source.

033. How people around the world earn

The World Bank publishes modelled estimates from the International Labour Organization (ILO). They show how the world’s workers earned their living in 2025:

• About 54% were wage and salaried employees.
• About 46% were self-employed.
• Employers, the self-employed who hire people to work for them, were about 4%.
• Own-account workers and contributing family workers, which the World Bank calls “vulnerable employment”, were about 42%.

Where you live changes the picture. Self-employment was about 11% of workers in high-income countries and about 79% in low-income countries; in the United States it was about 6%.

Much work is also informal. The ILO’s 2025 outlook says informal employment affects approximately three in five workers.

Our reading: self-employment is common worldwide, and far more common in low-income countries than in high-income ones. On its own, working for yourself is not a sign of wealth.

044. Why success stories mislead

We mostly see the survivors

Jerker Denrell made this point in a 2003 paper. The organisations we can observe “are the survivors of a selective process”. Books and the business press also tend to focus on successful firms, so the sample “usually undersamples failure”.

He argues that risky practices can then look linked to success among survivors, even when they are unrelated to performance across all organisations. Our page on survivorship bias shows the same trap in other settings.

What the official numbers say

The US Bureau of Labor Statistics tracks new business locations year by year:

• Of private-sector establishments that opened in the year to March 2019, 79.2% were still open a year later and 51.5% five years later.
• Of those that opened in the year to March 1994, 12.6% were still open in March 2025.

Two limits matter. The figures count establishments, so when a chain opens a new branch, it counts as a new opening. They come from unemployment insurance records, and BLS lists self-employed workers among the main exclusions.

Closed is not the same as failed

A 2003 paper by Brian Headd drew on two US Census Bureau data sources. One showed that about half of new employer firms survived beyond four years. Another, the Characteristics of Business Owners, showed that “about a third of closed businesses were successful at closure”. Many owners, he suggested, may have carried out a planned exit, sold a viable business or retired.

Can't confirm Nine out of ten new businesses fail. We found no official source for it. The official US figures count closures, not failures. About half of new locations were still open after five years; only after about three decades had close to nine in ten of the 1994 locations closed.

If a number about business is offered without its base, our guide to numbers that mislead has questions to ask.

055. Does money buy happiness?

This is a real scientific argument, and two researchers on opposite sides later worked on it together.

The 2010 study

Daniel Kahneman and Angus Deaton analysed more than 450,000 responses to a daily Gallup survey of 1,000 US residents. They separated two things: how people evaluate their life, and the emotional quality of their everyday experience.

• Life evaluation rose steadily with log income, that is, on a scale where each doubling of income counts the same.
• Emotional well-being rose too, but showed “no further progress beyond an annual income of ~$75,000”.

The 2021 study

Matthew Killingsworth drew on 1,725,994 in-the-moment reports from 33,391 employed US adults. Both kinds of well-being rose with log income, and he found no evidence of a plateau above $75,000 a year.

The two sides worked on it together

Killingsworth, Kahneman and Barbara Mellers then ran an adversarial collaboration to find an interpretation that fit both studies. They published it in 2023.

• Reanalysing Killingsworth’s data, they confirmed the flattening “only for the least happy people”.
• In the full paper they place it in the lowest 15 to 20% of the happiness distribution.
• In their words, “the suffering of the unhappy group diminishes as income increases up to 100k but very little beyond that”.
• Among happier people, happiness kept rising with log income, and even sped up in the happiest group.
• They concluded that the 2010 measures could not tell degrees of happiness apart because of a ceiling effect.

Not everyone agreed. In 2024 Ruben Arslan reanalysed Killingsworth’s aggregated data without assuming a threshold and found that “the evidence for a flattening is not strong”. The 2023 authors replied that the flattening “cannot be rejected” at the low end, while adding that this “does not prove” the 2010 pattern.

Both datasets come from the United States. The 2023 authors stress that the link is weak, even if statistically robust. In the 2021 data, the middle happiness score at a household income of $250,000 was about five points higher on a 100-point scale than at $15,000.

Partly Happiness stops rising above $75,000 a year. The 2023 reanalysis found a levelling-off only for the least happy 15 to 20%, at about $100,000, and a 2024 reanalysis said even that evidence is not strong. For happier people, happiness kept rising with income.

Is it the money, or something else?

In 2024 Julia Rohrer and Sebastian Wenz questioned the 2023 conclusions in the same journal. They argued that “cross-sectional associations only inform us about the diminishment of suffering in the absence of unobserved confounding”. In plainer words, a survey taken at one point in time can’t rule out that something else drives both income and happiness. They also argued that, without further assumptions such as rank invariance, the method does not refer to any specific group of people. They did not find rank invariance plausible.

The 2023 authors replied that causality “was not the point” of their paper and that their evidence was “strictly correlational”. They added that in the handful of studies that can test cause, “evidence suggests that having more money does, on average, lead to greater happiness”.

Partly More money causes more happiness. Both sides agree the 2023 finding is only an association. In a Swedish lottery study, large-prize winners reported higher life satisfaction for over a decade, with significantly smaller effects on happiness and mental health. A cash-transfer trial found gains in happiness, larger in lower-income countries. A British study found that medium-sized lottery wins went with better psychological health two years later. So the evidence leans towards more money raising well-being on average. The cash-transfer trial found gains in both life satisfaction and day-to-day feelings over three months; in the Swedish study, which looked 5 to 22 years after the win, the lasting rise was mainly in life satisfaction.

Our summary of The Psychology of Money looks at a related claim, that spending money to free up time makes people happier.

066. How to use it

None of this is financial advice. These are ways to read money claims, not money decisions.

• Ask for the missing failures. When someone tells you how they got rich, ask how many people tried the same thing, and where they are now.
• Read fail as closed until a source says otherwise. Then ask what happened to the owners, not only the shop.
• Ask which happiness. In the 2010 study, how do I rate my life? and how do my days feel? were measured separately, and they behaved differently.
• Treat averages as averages. These are findings about groups in one country, not a target income for any one person.
• Check your own country. The World Bank’s indicator pages let you look up how people earn in most countries.
• Look for tests, not testimonials. Our summary of The Lean Startup looks at what trials found when founders treated their plans as ideas to test.

07Quick answers

08Use this

09Quick quiz

How we checked this · Last checked October 2026

• We read the Greek and Latin dictionaries on Perseus, Pliny in Bostock and Riley's translation, and Merriam-Webster for the English words, including its note on the salt-money story.
• The labour figures are the World Bank's modelled ILO estimates for 2025, plus the ILO's 2025 outlook.
• The BLS survival table, its question-and-answer pages, the Headd abstract and Merriam-Webster block scripts, so we read them in a browser.
• For money and happiness we read the 2010, 2021 and 2023 papers, the two 2024 critiques and the authors' reply, and the abstracts of three studies that can test cause, two lottery studies and a cash-transfer trial.
• We found no official source for the claim that nine in ten new businesses fail, and none for Roman soldiers being paid in salt.

How we check every article →

10Sources