A study of 202,898 adults across 22 countries found that three widely used questions about wellbeing did not produce interchangeable pictures of a good life. Direct questions about life satisfaction and happiness tracked each other closely. The familiar request to place your life on a ladder from “worst possible” to “best possible” leaned more strongly toward wealth and material position.
The paper, led by Tim Lomas and published in Scientific Reports in 2026, is not evidence that one measure is false and another true. It shows that the wording of a broad question changes which part of wellbeing comes into focus.
This is one study, not settled consensus. The analysis is large and cross-national, but it is based on the first, cross-sectional wave of a longer project and uses one question for each measure.
What I find useful here is not a new recipe for wellbeing. It is evidence that the measuring instrument helps decide which version of wellbeing becomes visible.
The same people answered three different questions
The data came from the Global Flourishing Study analysis published in Scientific Reports. Gallup collected nationally representative samples within Argentina, Australia, Brazil, Egypt, Germany, Hong Kong, India, Indonesia, Israel, Japan, Kenya, Mexico, Nigeria, the Philippines, Poland, South Africa, Spain, Sweden, Tanzania, Turkey, the United Kingdom and the United States. Most of the first-wave data were collected in 2023.
Mainland China is part of the wider Global Flourishing Study, but its data were not available for this analysis. That reduced the sample from 207,919 to 202,898.
Participants first saw the Cantril Ladder. It asks them to imagine steps numbered zero to ten, with the worst possible life at the bottom and the best possible life at the top, then choose where they currently stand. They were also asked how satisfied they were with life as a whole, and how happy or unhappy they usually felt, using the same zero-to-ten range.
These sound like three versions of the same question. Statistically, they overlapped, but not completely. Life satisfaction and happiness correlated at 0.69. The ladder correlated at 0.57 with each of the other two measures.
The ladder was much more closely tied to national wealth
The clearest separation appeared in a post-hoc comparison with gross domestic product per person. Across the 22 countries, the Cantril Ladder had a correlation of 0.58 with GDP per capita. The corresponding correlations were 0.05 for life satisfaction and minus 0.08 for happiness.
Those numbers need caution. Twenty-two countries make a small sample for a country-level correlation, and correlation does not establish that wealth caused the responses. The analysis was also conducted after the main plan rather than specified in the preregistration.
Even so, the contrast is larger than a rounding difference. When the researchers grouped countries commonly described as Western, educated, industrialized, rich and democratic, their average ladder score was 6.60, compared with 5.81 elsewhere. Average life satisfaction was almost identical between the two broad groups, at 7.03 and 7.02.
The categories themselves flatten large differences between countries, a problem the authors acknowledge. Indonesia ranked high on life satisfaction and happiness while Turkey ranked low, though both sat in the same broad non-Western grouping.
Earlier experiments show what the ladder metaphor brings to mind
A separate 2024 experiment with 1,581 adults in the United Kingdom helps explain why the ladder might behave differently. August Håkan Nilsson and colleagues randomly assigned participants to variations of the question, then asked them to describe what the top of the scale meant.
With the original ladder wording, 17.3 percent of the words people used fell into the researchers’ power and money dictionaries. Removing the ladder reduced that share by six percentage points. Removing both the ladder and its top-to-bottom framing reduced it by 10.3 points. A harmony scale drew much more language about relationships, balance and peace.
The experiment does not show that everybody interprets the ladder materially. Most of the words used in the original condition were not about power or money. Its sample was British, recruited online, and the written responses were brief. What it establishes is that a visual metaphor presented at random can shift what people consider when they answer.
That finding fits the 2026 result. In the newer dataset, life satisfaction and happiness both had average correlations of 0.45 with the study’s other flourishing items, while the ladder averaged 0.38. The ladder came closer on financial and material security, suggesting that its difference may be informative rather than merely noise.
Most variation still existed within countries
It would be easy to read the findings as another national happiness ranking. The more useful result points in the opposite direction.
Differences between country averages accounted for an estimated 9.9 percent of variation in ladder scores, 7.7 percent in life satisfaction and 5.6 percent in happiness. More than 90 percent of the variation on each measure existed within countries.
Across the combined data, people who were retired, married or more educated tended to report higher evaluative wellbeing than comparison groups. Those unemployed and looking for work reported lower scores. Scores were lowest at ages 40 to 49 and highest among people aged 80 or older.
But the country-level exceptions were substantial. Retirees did not have the highest wellbeing on every measure in 11 countries, and they scored below unemployed people in Kenya and Tanzania. Age had no significant association with any of the three outcomes in Mexico and Spain.
These are associations, not instructions for how to live. Employment, marriage, education and age are tangled with income, health, public policy, selection effects and local culture. A cross-sectional survey cannot separate those pathways.
Life satisfaction may be the best single item, but one item is not enough
The authors tentatively recommend life satisfaction when a survey has room for only one evaluative wellbeing question. It correlated strongly with the wider flourishing domains and avoids the ambiguity of asking whether someone usually feels happy. If there is room for a second item, they recommend adding the ladder because it appears to capture different terrain.
That recommendation aligns with the OECD’s updated guidance on subjective wellbeing, which separates life evaluation, emotional states, and meaning or purpose instead of treating wellbeing as one undivided score.
There is a practical complication. Long-running surveys should not casually replace a question, even with a potentially better one, because doing so can break the historical series. The World Happiness Report has used ladder data for years. Consistency across time has value, provided readers understand what the measure emphasizes.
ScienceBlog has previously covered what an eight-decade longitudinal study can reveal about happiness. This paper offers a useful counterpoint: before interpreting a trend that lasts for decades, it matters which question was preserved for all that time.
The next waves can separate age from history
The first wave has several limits beyond its cross-sectional design. Childhood conditions were remembered in adulthood and are vulnerable to recall bias. Each wellbeing construct was represented by a single item. The three questions also appeared close together in the survey, which could prompt people to make their answers more different than they otherwise would.
The Global Flourishing Study is designed to follow participants annually through at least 2027, with its data made available through the Center for Open Science. Those repeated observations should help distinguish an age pattern from a cohort pattern, and a national difference from a temporary moment.
They will not remove the measurement problem. They will make it possible to see whether the three questions change together when the same people’s lives change, which is the comparison this first wave cannot yet provide.