My best guess is that Eric is basically right. In fact, I'd guess lower returns for the highest risk investments come from enough investors being risk-loving in relative wealth; they are willing to lose out on average for a chance to gain the very most. However, even if Eric is eventually proven very clearly right, I'm not optimistic that he will get much credit or gain from it.
The idea is that investors seek status. Since status is relative, investment strategy can be quite different from what would be expected under rational finance theory. For example, you would not necessarily diversify.
One of the things I dislike about stock-market contests for students is that they encourage this sort of thinking. I cannot imagine that you win the contest by choosing the optimal mean-variance trade-off. Instead, you find a way to take a strong bet on one type of outcome (say, a sharp rise in oil prices) and hope that the bet works.
It seems as though Hanson and Falkenstein see real-world investors using the contest mentality.