Posted by thatoneengineer 12/16/2025
Aren't all of these types of things (unhappiest day of the year, best day to be born on, age that we're happiest etc) clearly pseudo-scientific/scientistic babble - and brands can then just use them to sell the Scandi (or whatever) lifestyle. Nobody who believes this is going to be swayed by your anaylsis. :)
In such an environment it's vital to know if the methodology for measuring happiness is good or bunk.
But that just means the Cantril ladder is a good outcome variable — the WHR is in fact profoundly flawed, but the important flaws are in the predictors they use to “explain” (their word) the outcomes. They’re hand-picked, they’re over a decade old which is well before the majority of their own data was collected, they’re not even consistent with the report itself, and when they talk about them in public (I was at the 2025 launch party), they don’t even take them seriously, as if they know it’s not meaningful — and yet they continue to be the single largest data product of the report, every single year.
And this is critical. Who’s #1 is always in the headlines but Why is far more important than Rank. We don’t really care who has the best life — we want to know how we can get a better life. Yet most of the predictable conversation — here, but also literally on stage at Gallup — is just total speculation about the real answer, while sitting in front of 20 years of data. This is insane.
Which is why I’ve spent three years building a better model, starting from a base of 180x more variables, and using objective methods of computational variable selection instead of just deciding what I think should make people happy — because that’s self-evidently just inexcusably bad science. The result is measurably more accurate than the WHR. White paper is here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5655570
Tl:dr; Basic Needs, (Local and Global) Social Support, and (Local and Global) Self-Determination describe almost all of the findings, but many of the specific variables that emerge as the strongest predictors are things like LGBTQ+ social acceptance, women in white collar jobs, and meaningful, democratically accessible political power. Which just aren’t in the WHR model. The lessons to take, and the direction it points, are just in a profoundly different direction.
This is the real flaw of the WHR — it doesn’t actually show us how to make the world better.
Footnote: The Cantril ladder has now been used for literally 60 years, and new major studies continue to choose it as their outcome measure, because 60 years of research have demonstrated it is stable, meaningful, intuitive, and consistently understood across languages, cultures, and geographies. Plus it’s 1. self-reported, 2. all-encompassing, 3. single-scale, and 4. quantitative, all of which are unavoidable properties of a usable outcome, so even if the wording changes somewhat, any worthwhile question is going to look, basically, like it. And yes the tangled use of “happiness” vs “satisfaction” is stupid, misleading, and inconsistent, but when you just accept that one is the correct version and one is the PR version, you eventually get over it.
I read this critique when it came out, although you can really stop at the part where he claims that it's flawed because it's self-reported. This just totally, fundamentally, misses the point, and value, of the study. Do you think you should decide how good your life is? Or do you think I should decide how good your life is? Mounk appears to think he should be the one deciding (which is what you’re doing when you manufacture an “objective” version, rather than believing the provided answer). This is the deeper, and fatal, problem with his complaints. (The critique of the Ladder as being biased towards fame and fortune sounds important, until you actually model satisfaction and find that those variables just aren’t the dominant predictors.)
But that just means the Cantril ladder is a good outcome variable — the WHR is in fact profoundly flawed, but the important flaws are in the predictors they use to “explain” (their word) the outcomes. They’re hand-picked, they’re over a decade old which is well before the majority of their own data was collected, they’re not even consistent with the report itself, and when they talk about them in public (I was at the 2025 launch party), they don’t even take them seriously, as if they know it’s not meaningful — and yet they continue to be the single largest data product of the report, every single year.
And this is critical. Who’s #1 is always in the headlines but Why is far more important than Rank. We don’t really care who has the best life — we want to know how we can get a better life. Yet most of the predictable conversation — here, but also literally on stage at Gallup — is just total speculation about the real answer, while sitting in front of over 20 years of data. This is insane.
Which is why I’ve spent three years building a better model, starting from a base of 180x more variables, and using error-driven methods of computational variable selection instead of just deciding what I think should make people happy — because that’s self-evidently just inexcusably bad science. The result is measurably more accurate than the WHR. White paper is here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5655570
Tl:dr; Basic Needs, (Local and Global) Social Support, and (Local and Global) Self-Determination describe almost all of the findings, but many of the specific variables that emerge as the strongest predictors are things like LGBTQ+ social acceptance, women in white collar jobs, and meaningful, democratically accessible political power. Which just aren’t in the WHR model. The lessons to take, and the direction it points, are just in a profoundly different direction.
This is the real flaw of the WHR — it doesn’t actually show us how to make the world better.
Footnote, based on the conversation: The Cantril ladder has now been used for literally 60 years, and new major studies continue to choose it as their outcome measure, because 60 years of research have demonstrated it is stable, meaningful, intuitive, and consistently understood across languages, cultures, and geographies. Plus it’s 1. self-reported, 2. all-encompassing, 3. single-scale, and 4. quantitative, all of which are unavoidable properties of a usable outcome, so even if the wording changes somewhat, any worthwhile question is going to look, basically, like it. And yes the tangled use of “happiness” vs “satisfaction” is stupid, misleading, and inconsistent, but when you just accept that one is the correct version and one is the PR version, you eventually get over it.
The good news is that we don't need a perfect happiness report to think about the things various countries are either doing very well or very poorly and how our own lives might be changed if the place where we live did things differently. The World Happiness Reports gets attention year after year because it prompts that kind of thinking and there is value in that.