The incomes earned by elite athletes are often cited as examples in arguments to the effect that high incomes aren't a problem that need solving. If large numbers of people are willing - eager, even - to give small sums of money to watch Wilt Chamberlain play basketball, then on what grounds would anyone begrudge Chamberlain's salary or the system that generated it?
I was thinking about this while watching Bull Durham for the umpteenth time. Especially this part:
Just what is the difference between an athlete that ends up earning millions and one that never makes the big leagues? Luck is a part of it, but even if getting an extra hit in a given week may be a matter of chance, the law of large numbers says that you can't count on that kind of luck for long. But what this clip suggests is that the while the differences in skill between those playing in Yankee Stadium and those in the minor leagues may be small - one hit a week - they are at least measurable, and that's what counts.
To the extent that professional athletes' salaries can be explained by the tournament model, then those high salaries aren't so much a measure of the marginal product of top performers as an incentive dangled before everyone in the industry. Curiously enough, this is also the story that is often told to explain why CEOs' salaries are huge multiples of other executives.
So maybe we shouldn't be making the distinction between the millions that athletes earn and the millions earned by CEOs.