Mankiw, a professor of economics at Harvard, has written an opinion piece for the New York Times, on the topic of whether the rich deserve their large earnings. I had a tiny problem after reading through his piece, and that is the question of how he defines that pesky verb "to deserve."
This matters quite a bit. Does Mankiw mean that if we taxed the rich more, for example, something really bad would happen in the various markets they rule over? That the large earnings of the super-rich, say, are a necessary condition for the rest of us to have nice things at all?
Or does Mankiw mean that the high earnings of the top 0.01 percent, say, are deserved in some moral or ethical sense? And if this is the interpretation, does it have the corollary that the poor deserve their poverty?
Or does Mankiw mean something else, altogether, given that at the beginning of the piece he appeals to how people in general might feel about the question whether the rich are deserving or not:
In 2012, the actor Robert Downey Jr., played the role of Tony Stark, a.k.a. Iron Man, in “The Avengers.” For his work in that single film, Mr. Downey was paid an astounding $50 million.
Does that fact make you mad? Does his compensation strike you as a great injustice? Does it make you want to take to the streets in protest? These questions go to the heart of the debate over economic inequality, to which President Obama has recently been drawing attention.
Certainly, $50 million is a lot of money. The typical American would have to work for about 1,000 years in order to earn that much.
That sum puts Mr. Downey in the top ranks of American earners. Anything more than about $400,000 a year puts you in the much-talked-about 1 percent. If you earn more than about $10 million, you are in the top 1 percent of the top 1 percent. Mr. Downey makes it easily.
Yet, somehow, when I talk to people about it, most are not appalled by his income. Why?
One reason seems to be that they understand how he earned it. “The Avengers” was a blockbuster with worldwide box-office receipts of more than $1.5 billion.
Let's see what Mankiw's detailed argument are.
As shown in the above quote, he begins by referring to the "super-stars" phenomenon, the fact that current technologies allow the performances of a small number of very talented athletes, musicians and actors to capture gigantic markets, because of vast economies to scale, and that this results in humongous earnings for those superstars.
But Mankiw doesn't really discuss those economies to scale (the very low costs of letting large numbers of people enjoy watching or listening to the performance of one individual), or the possibility that those economies to scale can create monopolistic powers in markets, and once we have monopolistic powers, some parts of the earnings of the super-stars might accrue in the form of economic rents.
What are economic rents? Here's one definition:
In economics, economic rent is an analytic term for the portion of income paid to a factor of production in excess of its opportunity cost. Economic rent should not be confused with the more common term rent; a payment for the temporary use of a good or property.
Economic rent (which in production analysis is always seen as a cost of inputs) is affected by any production only minimally, if at all. Economic rent is a fact of natural or contrived exclusivity. For labor, economic rent could be created by the existence of guilds or labor unions (e.g., higher pay for workers, where political action creates a scarcity of such workers); for a produced commodity, economic rent may also be due to the legal ownership of a patent (a politically enforced right to the use of a process or ingredient); for operating licenses, it is the cost of permits and licenses that are politically controlled as to their number regardless of competence and willingness of those who wish to compete in the area being licensed; for most other production including agriculture, economic rent is due to scarcity of natural resources (i.e., land). When economic rent is privatized, the recipient of economic rent is referred to as a rentier.
By contrast, in economic theory, if there is no exclusivity and there is perfect competition, there are no economic rents, as competition drives prices down to their floor. Economic rent is different from other unearned and passive income, including contract rent. This distinction has important implications for public revenue and tax policy.  As long as there is sufficient accounting profit, governments can collect a portion of economic rent for the purpose of public finance. For example, economic rent can be collected by a government as royalties or extraction fees in the case of resources such as minerals and oil and gas.
Bolds are mine.
The simplified point is this: That super-stars earn large incomes does not prove that those large incomes couldn't be made somewhat smaller through taxation, without affecting the performance of the super-stars or the industries where super-stardom is common.
Mankiw ignores the possibility of economic rents in most of his opinion piece (nobody earns them, among the super-rich, though he does worry about rent-seeking behavior), so he doesn't really have to state whether he believes that economic rents are "deserved" or not, in the sense of economic outcomes being better when they are allowed vs. not.
Dean Baker makes a slightly different argument about this part of Mankiw's article (though it is still ultimately about how economic rents can be created). It has to do with the role of the government in enabling the huge earnings of super-stars:
His opening act is Robert Downey Jr. who apparently got $50 million for his starring role in a single movie. This is a great place to start. There's no doubt that Robert Downey is an extremely talented actor, but of course there have been many actors over the years who have put in great performances for much less money. How is that Downey could earn so much more than a great actor from the 50s, 60s, or 70s?
We could give a simple answer and say something like globalization and technology, but that would be at best half right. Certainly many more people will be able to see the films that Downey acts in than would have had the opportunity to see the stars from a half century ago, but that doesn't mean that Downey would get money from the broader exposure. In fact, a big part of the reason that Downey can collect huge paychecks is the extension and strengthening of copyrights. The United States has lengthened the period of copyrights from 28 years, with an option for a 28 year renewal, to 75 years in the 1976, and then to 95 years in 1998.
It also has stepped up copyright enforcement, imposing stiff fines on people who use the Internet to make unauthorized copies of copyrighted material. This is important, since the technology itself would let everyone quickly see Robert Downey Jr.'s new movies at no cost. It is only because of government intervention in the form of copyright monopolies that he is able to collect $50 million.
It is also worth noting that this intervention also has an indirect effect. If there was a large amount of high quality and recent material that everyone could obtain for free on the web (and show in theaters if they like), then no one would be willing to pay big bucks to see Downey's latest feature. So is Downey worth his $50 million, perhaps given the structure we have, but we could easily have a different structure which could quite possibly be a more efficient way to support and distribute creative work. (Here's my scheme.) FWIW, a similar story would apply to the writers and athletes in Mankiw's 1 percent defense.
Mankiw then moves from super athletes, actors and writers to something very different: CEOs. That smooth movement deserves our attention, because the reason why CEOs often earn a lot is not the same as the reason why the super-stars do, and my guess is that far fewer people would just accept that the high earnings of CEOs (including those golden parachutes and various deals for the failed CEOs) are obviously deserved*. The glide from the super-stars to the CEOs tends to infect the latter topic with what was said about the former topic.
Yet Mankiw's actual arguments differ between the two cases. Here's what he says about the CEOs:
Critics sometimes suggest that this high pay reflects the failure of corporate boards to do their job. Rather than representing shareholders, this argument goes, those boards are too cozy with the chief executives and pay them more than they are really worth.
Yet this argument fails to explain the behavior of closely held corporations. A private equity group with a controlling interest in a firm does not face this supposed principal-agent problem between shareholders and boards, and yet these closely held firms also pay their chief executives similarly high compensation. In light of this, the most natural explanation of high C.E.O. pay is that the value of a good C.E.O. is extraordinarily high.
That is hardly a surprise. A typical chief executive is overseeing billions of dollars of shareholder wealth as well as thousands of employees. The value of making the right decisions is tremendous. Just consider the role of Steve Jobs in the rise of Apple and its path-breaking products.
A similar case is the finance industry, where many hefty compensation packages can be found. There is no doubt that this sector plays a crucial economic role. Those who work in banking, venture capital and other financial firms are in charge of allocating the economy’s investment resources. They decide, in a decentralized and competitive way, which companies and industries will shrink and which will grow. It makes sense that a nation would allocate many of its most talented and thus highly compensated individuals to the task.
In addition, recent research establishes that those working in finance face particularly risky incomes. Greater risk requires greater reward.
Dean Baker is worth reading on this part, too. For example, he points out several of the problems in the above Mankiw quote:
Then we get to the CEOs who Mankiw tells us get high pay because of what they contribute to their companies and the economy. If this is the case, how do we explain CEO's of companies like Lehman, Bear Stearns, and AIG walking away with hundreds of millions of dollars even though they drove their firms into bankruptcy? When the CEO of Exxon-Mobil gets hundreds of millions because soaring worldwide oil prices sent Exxon's profits through the roof, do we really think the pay is a function of hard work? How do we explain the fact that CEOs of incredibly successful companies in Europe, Japan, and South Korea make on average around a tenth as much as our crew does?In short, one can argue against the assumptions that the CEOs are simply paid a lot because of their rare talents and that the markets for chief executives are nicely competitive.
Likewise, Mankiw's reference to "greater risk requiring a greater reward" leads to a study, not about risk in general, but about business cycle volatility of earnings, with an abstract which tells us this:
How sensitive are the earnings of top earners to business cycles? And, how does the business cycle sensitivity of top earners vary by industry? We use a confidential dataset on earnings histories of US males from the Social Security Administration. On average, individuals in the top 1% of the earnings distribution are slightly more cyclical than the population average. But there are large differences across sectors: Top earners in Finance, Insurance, and Real Estate (FIRE) and Construction face substantial business cycle volatility, whereas those in Services (who make up 40% of individuals in the top 1 percent) have earnings that are less cyclical than the average worker.
Bolds are mine.
The bolded sentence reminds us that a sizable percentage of those whose earnings are in the top one percent face less business cycle volatility than the average worker. The greater risk argument cannot be used as an explanation there.
Then there's this bit by Mankiw:
Those who work in banking, venture capital and other financial firms are in charge of allocating the economy’s investment resources. They decide, in a decentralized and competitive way, which companies and industries will shrink and which will grow.
Except when they perhaps don't carry out these tasks in a decentralized and competitive way. Matt Taibbi writes about the possible problems with that assumption:
Today, banks like Morgan Stanley, JPMorgan Chase and Goldman Sachs own oil tankers, run airports and control huge quantities of coal, natural gas, heating oil, electric power and precious metals. They likewise can now be found exerting direct control over the supply of a whole galaxy of raw materials crucial to world industry and to society in general, including everything from food products to metals like zinc, copper, tin, nickel and, most infamously thanks to a recent high-profile scandal, aluminum. And they're doing it not just here but abroad as well: In Denmark, thousands took to the streets in protest in recent weeks, vampire-squid banners in hand, when news came out that Goldman Sachs was about to buy a 19 percent stake in Dong Energy, a national electric provider. The furor inspired mass resignations of ministers from the government's ruling coalition, as the Danish public wondered how an American investment bank could possibly hold so much influence over the state energy grid.
*One quote about the opinions concerning the optimal distribution of wealth in the US suggests this:
The most astonishing piece of social science research I’ve seen in some time was published in late 2011 by two academics, Michael Norton of Harvard and Dan Ariely of Duke. They asked a sampling of Americans two basic questions: What do you think wealth distribution in the United States is today, and what wealth distribution do you think would be ideal? They then matched those two sets of numbers to the existing facts.
Respondents guessed that the top 20 percent of Americans owned just under 60 percent of the wealth, and the bottom 60 percent owned just more than 20 percent. Their ideal distribution, they said, would be for the top quintile to own only about 32 percent of the wealth, and the bottom three quintiles to own about 45 percent. The actual numbers: the top quintile owns more than 80 percent, while the bottom 60 percent owns around 5 percent. The results suggested that if Americans knew all this, the political space for a more aggressive left-populism would exist.
Added later: It's still worth reading Robert Solow's response to something similar Mankiw wrote earlier, as well as Mankiw's response to that.