2. More talented people tend to earn higher incomes.My guess is that the story here is far more complex than Mankiw wants to admit (perhaps even to himself), but in general not nearly so true as he thinks it is, or at least not in the same way.
We can start by dispensing with the circularity: rich people on the whole probably have a talent for getting rich. I say "probably" here to include cases of sheer accident, like The Inheritors. But setting them aside, my point is that if we want to make any sense out of this analysis we have to define "talent" as something other than "whatever got you here."
In fairness, I don't think Mankiw makes this error. He seems willing to concede that we can measure talent by something exogenous like, say, IQ tests, so let's play along. Specifically, my suspicion is that the relation between talent=IQ and high incomes is not nearly as linear or highly correlated as Mankiw thinks it is. Or rather: up to a point I suppose that IQ leads to higher income, but I suspect the point is fairly modest and well defined. Say, maybe 125-130.
What happens beyond 130? Two things. One is that beyond this point, extra intelligence doesn't really help all that much. For most high-income jobs, you probably have to read, to remember a bunch of stuff, for some perhaps to add and subtract. But the number of jobs that require really first-rate IQ is pretty modest (more on those jobs in a moment). And two, the flip side--for many high-income jobs, high intelligence probably gets in the way of good performance. It is very likely to associate with poor social skills, maybe with some degree of Asperger's Syndrome (if there is such a thing. If you are a lawyer, they probably keep you in the library writing briefs; if a doctor, maybe you are in a skill that does not require a bedside manner, like maybe pathology. But short of that, if you are really bright, there is a good chance you are in your mother's basement waiting for a guest spot on Wayne's World.
To test yourself on this, among presidents, whom do you prefer: the really bright guys like Woodrow Wilson or Jimmy Carter or Richard Nixon? Or the less bright but more stable and balanced like Franklin D. Roosevelt or Dwight D. Eisenhower or Ronald Reagan?
As an example for his own case, Mankiw deploys Krugman himself, a presumptively bright guy. But the choice of Krugman shows just how weak his argument is. Because the thing about Krugman is that he is a professor--and let's face it, academia is one place where poor social skills are actually rewarded. I'm a great Krugman admirer and I think he is right far more often than Greg Mankiw. But I've heard Krugman himself say that nobody should trust him with a government job and I couldn't agree more. Hell, we've already got ourselves into enough trouble with this guy.