Rational or Irrational: The Logic of Life or Predictably Irrational? (Author One-on-One: Dan Ariely)
The Logic of Life is a wonderful book. In prose that is witty and entertaining, Tim Harford uses his keen economist's perspective to peel away the layers from a variety of unconventional examples, revealing the unexpected rationalities that lie at the hearts of subjects such as large executive salaries, smoking, and even racism. His exploration of these subjects enables the reader to understand how economists think, and the logic that often lies beneath the surface.
Take for example the case of executive pay. As Harford explains, there are some interesting economic principles underlying what we might otherwise consider exuberant CEO salaries (for example the $800 million that Disney paid Michael Eisner over his 13 years of service as Disney's CEO). First, Harford argues that very high salaries are desirable because they cause the executives to have more skin in the game and, as a consequence, will care more about their company. Harford also points out that these high salaries are not only intended to motivate the person who is getting the high salary, as they are also intended to motivate the people below them who might pull all-nighters for years and work much harder in hopes of one day attaining that enviable position. (Harford proposes a clever test of this idea in which the CEO doesn't do any work, making his position even more enviable.) Finally, Harford also points to the rational selfishness of the CEOs, the difficulty of tracking their actions, and the mysteries of accounting as other ways in which CEOs further increase their salaries.
These are very interesting ideas with a compelling and fascinating logic (although they are not very well tested from an empirical perspective). In my mind this is economic theory at its best. Beautiful, elegant, requiring some convoluted logic, and shedding light on what we see in the world around us.
What should we do with these insights from economics? If we trust the rational economic perspective completely, we should take steps to pay CEOs more, maybe give them more vacation time or pay them even more with stock options. But here is the catch: in order to move from a perspective that sheds light onto the question of CEO pay to an action plan that we can adopt wholeheartedly, we must not only believe in these principles, but we must also believe that nothing else matters (or at least that nothing else would make a big difference). So, should we trust that our powers of logical reasoning could provide us with a complete understanding of the role of incentives? The field of behavioral economics suggests that we should not rush to make this assumption and that, in fact, there are many forces that influence us without our understanding.
Let me give you an example. Imagine that I described to you some puzzles of different sorts that required memory, concentration, thinking, and creativity. Suppose I then asked you to predict how well people would perform on these games if they were promised a payment that depended on their performance. I would also tell you that for some people the payment was equivalent to 1.5 days of their income, for others it was equivalent to 2 weeks of their income, and for a third group it was equivalent to 6 months of their income. Who would you predict would do the best? If you are anything like the real participants in our experiments, you would predict the third group--the one with the highest pay--would have the highest level of performance.
When we ran this experiment (in India where we could pay people this much without completely depleting our research accounts) we found that participants did the worst when they were offered the highest pay. It turns out that money is a double-edged sword: it is a motivator but it is also a stressor. To get a better insight about this idea, imagine that I offered you $10,000 to come up with a creative idea in the next 5 minutes. How focused would you be on the creativity task and how stressed would you be? How much time would you dedicate to coming up with an idea and how much time would you spend thinking about the money that you are not going to get?
What is the big idea here? First, it turns out that incentives are a bit more interesting and complex than standard economics would have us believe--sometimes we can pay more and get worse performance. Second, it appears that our intuitions and logic are sometimes inaccurate in their ability to predicting human behavior. What does this say about relying on standard economics and pure logic as a tool to guide the type of incentive mechanisms we create? It tells us that when we try to understand these mechanisms, and in particular when we make recommendations for how these mechanisms should be designed (which is something economists do often), we should take all the input we can into account, including our irrational characteristics. Standard economics and logic can help us create systems that are useful for perfectly rational people, but behavioral economics will help us design a better world for the rest of us.
Irrationally yours,
Dan
Stacey on April 16, 2008 at 09:45 AM
I'm sorry, but paying a CEO hundreds of times the average pay for the rest of the company is terrible. The real reason salaries are so inflated is competition between companies to hire these people, just like in pro sports. However, the rest of the working world does not have a union negotiating minimum salaries for them so only the CEO gets to reap ridiculous pay. That, plus they all get great golden parachutes to cash in when they can't do the job (and/or break the law trying to please Wall Street too much).
CEOs should be compensated well, but publicly traded companies should have boards that act responsibly with compensation. I am quite sure that Michael Eisner did not do $800 million better than someone else in his position could have.
Great book, Dan. Look forward to more of these posts.
Dave on April 16, 2008 at 10:47 AM
Stacey, there are some problems with your comment - it seems to be based on emotion (how appropriate for a Predictably Irrational blog!). First, you only looked at the demand side of the labor market. And sure enough, competition between companies does bid up the salaries of CEOs, if they are in limited supply. But there is also competition among potential CEOs for these jobs. The more qualified candidates (supply of potential CEOs) there are, the more downward pressure you would expect to see on CEO salaries. And there are certainly many business men and women waiting to fill the next CEO posts.
It's interesting that you used pro sports as an example, because many pro sports have maximum and minimum salary constraints which suggests that the top stars are likely to be underpaid (because the cap prevents bidding to exceed a certain level) and the bench warmers overpaid. In other words, there is already income redistribution in many professional sports.
It's also interesting that you have brought up the topic of unions. First of all, CEOs don't have unions negotiating for them, so I'm not sure why it's unfair that other workers do not have them either. Secondly, unions that demand higher compensation have encouraged US companies to offshore production. They hurt, rather than help in today's global marketplace.
In my opinion, what seems to be problematic with CEO compensation is that a) heavy stock option compensation can encourage too much risk taking by not internalizing the risks CEOs take and b) CEO compensation is decided by board members who tend to be CEOs of other companies and want to keep the market prices of CEOs high.
Certainly some CEOs are overpaid and didn't deserve the golden parachutes to which you refer. But just because someone makes a lot of money doesn't mean they do not deserve it.
Gabriel on April 16, 2008 at 11:19 AM
>> "If we trust the rational economic perspective completely, we should take steps to pay CEOs more [..]"
It does not follow, I think.
If everyone does their best, given circumstances and others' similar behavior, it's not obvious why the realized outcome would call for any policy or prescription. -- And even if some policy would be called-for, it's not clear why it would be this one.
Andy on April 16, 2008 at 11:29 AM
I agree with Gabriel. Maybe the current salaries for CEOs are efficient? Possibly paying them any more isn't worth the marginal cost in terms of improving worker productivity.
RobbL on April 16, 2008 at 05:57 PM
Dan,
I wonder if what you are saying is really a more general attack on the whole project of logic in Economics? I have to say that I have a lot of agreement with this idea.
It seems to me that the histories of all sorts of academic disciplines are littered with ideas that seemed logical, but turned out not to be true. That is the problem with logic....there are no true foundations to argue from and no true facts to argue with.
So we get all sorts of "paradoxes" that really just prove that reality is complex and simple logical models are just that...models.
When you try and translate them into the real world, all sorts of "illogical" things happen.
I like the story (I think the reality of this has been questioned, but it still seems logical :-) ) about the day care center that didn't like parents picking the kids up late so they added a late fee. This resulted in even more parents being late. It turned out (in economic terms) that this lowered the price of being late because it apparently lowered the social censure cost of being late once the center put a money price on lateness.
The interesting thing about this (always making the assumption that this is true or could be true) is that it destroys the supply and demand curve in standard economics. The general assumption is that we understand the units on the axis of the curve (generally it is assumed that money is key) but once you say that you don't know for sure what the units are, the model doesn't seem to have much use.
In the abstract the curve is perfectly valid, but in everyday application is not of any use because we don't know what to map to.
DerekT on April 18, 2008 at 02:03 AM
Dan
Your suggestion that the offer of 6 months salary wouldbe the "rational" choice is not, for me, rational. Perhaps the "rational" response is that size of offer indicates that, in order to win it we think the challenge may require more effort/knowledge or whatever than we might have? ie the more "reasonable" the offer, the more likely it would be that we could win it?
NB Haven't bought your book - yet but the reviews make it interesting, so it's on the list (have to finish Tim's book first!).