What the HECK is `earnings penalty`

And it appears to be implying that while the percentage return on education over time has actually been increasing, the return for artists actually capped out in the 90's, and then declined recently. Course, it only has one date point per decade, has no units, and little in the way of description, so its a bit light on data. The apparent art decline could be a statistical anomaly depending on how they gathered it.

On the penalty, I don't know the context, but it might also be talking about the advancement of professionals in school vs going out and working. Some professions increase more rapidly with education, while others do so with field work. Hard to tell the way you presented it.

Jam it back in, in the dark.