Thanks for the link. That's an important point. I think part of the problem is the pace of change. That's both the reason for the bad data and the time lag iss

Thanks for the link. That's an important point. I think part of the problem is the pace of change. That's both the reason for the bad data and the time lag issue. It's just impossible to make a decent guess at the necessity of acquirable skills on a useful timescale. Even technical skills that have a fantastic outlook from todays vantage point… Our visibility is not really that far, 15-20 years at a stretch. For example, it seems like a good bet that for the next 15-20 the skills that allow you to write programs for a living will still be marketable, assuming a non trivial amount of re-skilling along the way. past that, who knows. For a 20 year old, that's one third of a career. Take the skills required to be employed in the enormous and affluent financial sector. Accounting, finance, data analysis, math. Can you imagine that over the next 15-20 years something happens (bitcoin?) that turns this business on its head and makes a big chunk of its employees obsolete? For the opposite, can you imagine that 15-20 years from now gene-tech has created industries that demand geneticists the way programmers are now in demand? Pace of change is itself a meaningful change.