That’s why I hate these artificially drawn “divisions” between groups of fields. Content-wise, you get weirdness going on - anthropology has very humanistic aspects to it and in some ways has more in common with humanities; history has very social science aspects to it; and a particular given anthropologist might function/do scholarship more like a humanist than a social scientist. Psychology sort of straddles the social and natural sciences, particularly in some subfields; biology, especially some subfields, can have a lot of social sciency aspects to it.
But the other thing is career-wise splitting them up this way implies the STEM PhDs always have an easier time getting non-academic positions than non-STEM PhDs in the social sciences and humanities. It’s really more dependent upon your field, and drilling deeper than that, it depends on the kind of research that you do within your field. For example, quantitative psychology is a branch of psychology that deals with the development of statistical methodologies for psychological research; a quantitative psychology PhD (or a heavily quantitative political scientist, for example) probably has better non-academic job prospects in the world of data science and quantitative finance than a biologist or a chemist who doesn’t use statistics - or even a physicist who does not (I recently met a physicist who does not use any statistics in her work, and she admits to not knowing much about it).
The other question is “for what?” Most marketing jobs, for example, ask for social science majors. I think a PhD in anthropology or sociology has a better shot at those kinds of jobs than a PhD in mathematics or engineering. I say this to mean that I don’t think that STEM PhDs have more applications in the non-academic world; rather, I think STEM PhDs have more applications in the non-academic world that don’t require a complete re-imagining of the work that they do and the value that they offer. It doesn’t take much imagination for a PhD in electrical engineering to go from doing research in academia to doing research for AT&T or Apple, or to go work as a professional engineer; it doesn’t require struggle for a PhD in math to imagine doing quantitative finance or other mathematical modeling positions outside of academia.
It’s harder for anthropologists and historians and English literature scholars to imagine counterparts in the non-academic world - even though I very much do know of people who are not academics who are paid to essentially be anthropologists, historians, and literature scholars in the private or public non-academic sector, as well as people who use the skills from those PhDs to do different things.
With that said, economics PhD holders have good prospects for two reasons. One, their quantitative expertise and knowledge of something that is very remunerative in the non-academic world (economics) does make them in demand, both at private companies (banks, insurance firms, investment/hedge funds, etc.) and public sector type jobs (World Bank, IMF, UNESCO, whatever). Two, the very fact that their skills are so in-demand outside of academia - and that they are so well-paid there - means that the competition for actual academic jobs is lower than it is in other social science fields. So it’s also a bit easier for them to get jobs within academia.
The same is true of business PhDs on a much larger scale. There’s actually a shortage of business PhDs to teach at U.S. business schools, particularly in the fields of accounting and finance. So a PhD in accounting, if they wanted to be an academic and had a decent/average publication/research record, could probably find a great academic job. But yes, there’s tremendous value for business PhDs outside of academia - although of course that’s going to vary based on subfield (the PhDs in finance will very likely come out way better compensated than the marketing PhDs, and both may have an easier time finding employment than management PhDs). It’s also going to depend a lot on their experience.