Economists Do It With Models

Warning: “graphic” content…

Bookmark and Share
The One Percent, Genetic Lotteries, And Tournament Incentives…

June 26th, 2013 · 3 Comments
Uncategorizable

In case you haven’t been following, it seems like every economics writer on the interwebs has gotten his panties in a twist over Greg Mankiw’s forthcoming article in the Journal of Economic Perspectives entitled “Defending the One Percent.” (Mankiw’s thesis is pretty much exactly what you would expect given the title.) My gut reaction is to say that most people with at least one foot in reality will disagree with most of the unsubstantiated assertions in the paper (like the one that, despite everything that Joe Stiglitz has said over the last few years, rent seeking isn’t really a substantial thing, or perhaps that fancy-pants education is basically the same as the average education in the U.S.) but will also be frustrated that Mankiw refuses to make the (far more reasonable, in my view) assumption that, on average, someone like me would get more use out of an extra dollar than someone like Bill Gates. That said, the paper provides an interesting discussion on utilitarianism and political philosophy. I won’t rehash the main objections, which have been, sometimes colorfully, written about elsewhere. If you are interested in such commentary, here are a few examples:

I think such commenters have covered most of the possible ground on the issue, but I do feel that they left out two pertinent issues, one that goes against Mankiw’s argument and one that supports some of his premises to a degree.

The first point concerns the degree to which “equality of opportunity,” as Mankiw and many others define it, guarantees an efficient and fair distribution of income in society, and, by extension, begs the question of whether “equality of opportunity” has been properly defined. Sooo…let’s talk about models. (Sidenote: I’ve been waiting soooo long to be able to say that.) Fashion models get paid a lot of money largely due to the fact that they are in limited supply. (If you live somewhere where all of the women look like Gisele or Kate Moss, let me know so I can share with readers as a public service announcement…though you may not want the competition!) Said models are, in part, collecting rents because they won a genetic lottery. In other words, it’s not like they are going to get uglier if you don’t pay them as much. That said, there is an effort component to the human capital required to be a model- in part, models are getting paid to take care of themselves, have their salad dressing on the side, etc. Maybe the cost of effort for models is lower than for the general population- Kate Moss famously stated that “nothing tastes as good as skinny feels,” and I often wonder whether her taste buds are not as developed as mine or whether her returns to skinny are higher than mine, since the former would in fact speak to individual differences in the cost of effort. Either way, models do require the proper incentives to induce the effort required to properly complement their genetic gifts, but it’s unclear how their proper incentives compare to either their actual compensation or what my incentives would have to be in order to elicit the same behavior and/or outcome. (I would not be surprised if no amount of money could prevent me from eating truffle fries.)

For these reasons, I don’t think it’s particularly controversial to say that it would be efficient to compensate models more for what they can control versus what they were endowed with, or even that it’s fair to compensate models less since much of their human capital comes naturally to them. (Note, however, that you also need to compensate enough so that models become models and not other things.) Now let’s bring the context back to the children of the smart parents that Mankiw refers to, since, according to him, they were also winners of a genetic lottery, albeit a less obvious one. Are these children going to become dumber if they are not compensated adequately? Potential substance abuse issues aside, probably not. Do these children have lower cost of effort in acquiring skills and doing a lot of jobs? Most likely. At the risk of repeating Mankiw’s error of personal anecdote (and sounding like a giant brat, which is certainly not my intention), I can’t help but ponder how it at least seems like it’s less painful for me to learn nerdy stuff than it is for a lot of other people. As a result, would I be willing to teach economics (and acquire the skills in order to be able to do so) for less compensation than the average person? Probably, since achieving the required level of productivity is easier for me than for others. Would it be efficient to pay me less because of this? In most circumstances, yes. Would it be fair to pay me less because of this? I don’t know, but I can’t comfortably rule out the possibility. (If you are my employer, please forget that I wrote those last few sentences.)

To some degree, wages based on the supply and demand of labor reflect this consideration, since, for instance, people like me are going to supply labor in the market for economics lecturers at a lower wage than a lot of other people because of our relative ease of doing the job. But when the supply of cheap people like me is limited compared to demand, the market equilibrium wage is going to benefit me financially due to the fact that what is easy for me is harder for other people, even though much of what makes things easy for me was supposedly bestowed on me at birth. For this reason as well as those previously stated by others, I can entirely understand how people don’t find a philosophy of “income inequality is a-ok because it represents differences in productivity” or, more accurately, “large income differentials are necessary in order to elicit optimal productivity in society” to be particularly compelling. (In related news, parents’ socioeconomic status doesn’t seem to be as large a determinant of children’s cognitive ability as Mankiw suggests, which, strangely enough, might actually help his argument.) I’m not claiming to know what the correct policy implication is here, but hopefully the discussion at least highlights the narrowness of the typical definition of “equal opportunity”- does everyone really have equal opportunity when the things that are most valued by the market are easier for some people than for others? This sort of inequality doesn’t necessarily lead to an inefficient allocation of resources, but, as Mankiw himself notes, what is efficient isn’t necessarily fair.

Another issue that people seem to be picking on in Mankiw’s paper is the assertion that an individual’s compensation actually reflects his or her productivity in the first place. While Mankiw’s assumption may be incorrect, it need not be the case that compensation matches with productivity in order for compensation to provide efficient incentives for workers and for society. For example, one common economic explanation for CEO pay that exceeds marginal product is that the organization is structuring compensation and incentives in a relative, or tournament, sense, and the CEO happens to be the winner of said tournament. Granted, paying “extra” to the CEO because he won the corporate leapfrog game isn’t great from the perspective of income inequality, but it’s possible that it works as well to motivate an organization as a system where each worker is paid his marginal product. (This is because the possibility of eventually winning the CEO tournament and getting the big paycheck serves as an indirect incentive to those who are still working their way up the corporate ladder.) In fact, tournament compensation can be superior to paying each worker his marginal product when marginal products are hard to observe but relative performance is fairly clear. In other words, it’s entirely possible that taxing a CEO at a higher rate may have an effect on society overall even if it doesn’t change the behavior of the current CEOs.

Lastly, on a personal note…I have a story that very much parallels Mankiw’s in a number of ways, even down to parental ethnicity. (My parents weren’t immigrants to the U.S., but they did move when I was little out of the highest-poverty major city in the U.S. Hey Detroit, you’ve got nothing on Youngstown.) In general, I am very tempted to, like Mankiw, make the “if things worked for me” argument, but, ironically enough, it is Greg Mankiw’s own textbook that taught me to resist the urge. When I was one of Greg’s TAs, I had to teach a section on income distribution, which of course required me to pay attention to the statistics on income distribution. I would never ask them directly, but I distinctly remember what my parents’ reported income was on my college financial aid forms, and I was a bit taken aback to learn that what I had considered to be an average or slightly below-average financial environment was actually approaching the top quintile in terms of income. (I am not alone- people are pretty bad at guessing where they fit in the income distribution. For context, my parents teach/taught high school and don’t spend extravagantly, so I now conclude that it’s pretty easy to confuse “comfortable but not extravagant” with “average,” probably because I feel that that *should* be the average experience.) Knowing this, I became much more careful about extrapolating from my own experience, both because most families don’t have the resources that mine had and because not all families make education and achievement the priority that mine did. (In related news, my mother would like me to mention that she is not a tiger mom, since that is often incorrectly assumed. Let’s call her an enabler instead- I’m sure she’ll love that.) Knowing what I know now, my experience instead highlights the fact that, until kids can choose their parents, even equality of income won’t necessarily lead to equality of opportunity. Luckily, we’re starting to see data on what I will call the marginal product of good guidance counselors.

Update: Noah Smith offers his version of my first point. He addresses the fairness aspect rather than the efficiency aspect, but I think his post makes that part of the argument in a very clear and effective way.

Tags: Uncategorizable

3 responses so far ↓

  • 1 Frederic Mari // Jun 27, 2013 at 9:37 am

    I find the tournament hypothesis intriguing but unlikely to actually be the justification of high CEO pay.

    It will, early enough, be obvious to a lot of workers that they are not on the CEO track.

    If paying them less (than MP or whatever) in order to pay the CEO more leads them to being disgruntled (actively disengaged as one article recently described saboteurs) then high CEO pay is actually counter-productive.

    Outside of the world of economic model, I find it easier to simply observe how CEO pay is set. As it happens, it is set by themselves , their friends and enablers and without much control from the markets or the shareholders. And unions are now unable to protest on top.

    In those conditions, I find it not hard at all to believe they rationally decide to loot their corporations. It simply makes sense.

  • 2 uair01 // Jun 29, 2013 at 8:11 am

    Funny to read this on another weblog:

    It was interesting to see what textbooks people use and what they think about them. Greg Mankiw’s Economics is a market leader in many countries, but none of the participants yesterday like it.

    http://www.enlightenmenteconomics.com/blog/index.php/2013/06/teaching-economics/?utm_source=rss&utm_medium=rss&utm_campaign=teaching-economics

  • 3 Jeff // Jul 2, 2013 at 11:16 pm

    Wow! A challenge! Find a criticism not already lobbed. By the way, if you read my FB response on the same subject, do you forgive me yet? (Why is it that I always have to ask that?)

    Anyway, a new and different criticism.

    One thing you learn in the world of software is that the more expensive a product is, the more bugs it has, not even total, but per line of code. This is not because people writing expensive software are any less skilled than people writing cheap software, although that has been known to happen. It is simply a matter of eyeballs, which before now was only popularly known as the creepiest investment craze since two-lip mania (ok, wait; sources are informing me that was actually about Tulips. Never mind…). Anyway, the concept is, the more eyeballs see a piece of software, the more bugs are removed from it even if the dollars per line of code is the same. This is true both from a simple literal perspective and from an incentives perspective (think economies of scale in customer support, which is often more expensive than writing the code in the first place). If you carry this idea to its logical conclusion, there are a lot of pieces of technology we wouldn’t have without a middle class even if you assume that physical capital accumulation could somehow occur at the same rate in a highly unequal society, which it obviously can’t.

    This thought experiment demonstrates why there are not only quantitative, but also significant qualitative differences between an “equal enough to grow” economy and an “unequal enough to shrink” economy. The rich need the poor so they can get X-Boxes.

Leave a Comment