> in 2026 there is still a striking college income premium, statistically (it's over 50%).

We can make this real simple: 64% of the population have a degree. The median income is unchanged since before <50% of the population had a degree. If individuals gained a premium, we'd have to have seen an increase in the median income after >50% of the population attained a degree. That is a mathematical necessity. But it didn't happen... There is no college premium, plain and simple.

While incomes are stagnant for nearly all individuals, college degree or not, I suspect what you are trying to point out is that the so-called top 1% has seen a dramatic rise in income and that most in the top 1% have a degree. Therefore, the combined income of all degree holders has risen at a rate faster than the remainder of the population, even if most degree holders have seen no change. That is definitely true, but a long way from your claim that education influences lifetime earnings. Given that only a small subset of all degree holders have found what could be called a premium, it is unlikely that the premium is influenced by college; at least not in a consistent or reproducible way to make it a useful claim.