Luck in Economic Analysis

I recently read a couple of articles about the role of luck applied in economic analysis. The first was by Moshe Levy. Levy’s article noted research he had done that showed the effect of luck on executive compenation (Levy, M. (2016). 90 Cents of Every’Pay-For-Performance’Dollar Are Paid for Luck. Browser Download This Paper) and earlier research examining the link between luck and wealth inequality (Levy, M., & Levy, H. (2003). Investment talent and the Pareto wealth distribution: Theoretical and experimental analysis. Review of Economics and Statistics, 85(3), 709-725.). The other article by Bob Henderson was an interview with Robert Frank also about the role of luck in being successful. Frank recently published a book (Frank, R. H. (2016). Success and luck: Good fortune and the myth of meritocracy. Princeton University Press) in which he discusses the connection between luck and success and how they relate to the notion of a meritocracy.

In a sense, both articles were attacking the notion of self-driven success. Levy looked at so-called “pay for performance” in executive compensation. The context is a response to the observation that the average CEO earns 480 times the salary of the average worker. One argument for this disparity is that the business must compensate “rare managerial talent. Levy (2016, p. 1).  Building on an earlier work, Bertrand, M. & Mullainathan, S. Are CEOs rewarded for luck? The ones without principals are. Quarterly Journal of Economics 116, 901-932 (2001), Levy’s paper seeks to quantify the proportion of “pay-for-performance” that is really attributable to luck and calculates a figure of 90%.

Levy’s earlier paper, Levy and Levy (2003) employed simulations in attempt to explain why the distribution of wealth fits a power function distribution. Their results rely upon two details. First, that the distribution of wealth fits a power function distribution, and second, that their simulation shows that an approximate power function distribution of wealth only emerges from an investment market if there is no distribution of talent, i.e., luck dominates investment outcomes. One additional interesting finding is that it does not take very long for the distribution of wealth to fit a power function distribution (years, as opposed to generations).

While Levy shows that pronounced impacts from luck are readily observable, Henderson’s interview with Frank focuses on how people perceive the impact of luck within their own lives. Frank notes that experimental evidence has shown that people do not tend to appreciate the role of luck attributable to success. In his book, Frank goes through the likelihood of good luck occurring, how good luck impacts individuals, and provides some simulation results that explain why very successful people are always very lucky people (or, rather, that their immense success is largely attributable to their good luck). Frank also notes in the interview that it is possible to get people to recognize the effect of luck on their successes.

A few years ago I was teaching two classes of introduction to economics (micro and macro combined) and at the end of the term I decided to talk about the distribution of wealth and wealth inequality. My students were nearly unanimous in their belief that people are largely rewarded for their marginal product and that talent and effort largely drives income and wealth. As an exercise I had both classes play 5-card draw poker. I brought in fake money and decks of playing cards. I was initially surprised by the number of students who had no idea how to play poker. For the exercise, the class was split into groups of 5 students, each student was given an equal sum of fake money and each group was given a deck of playing cards. Groups played three hands of 5-card draw poker and then the students were re-grouped. Students played in three groups. Students were awarded extra credit points depending on the final amount of fake money they possessed, so they did have some incentive.

At then end of class, students were instructed to count and report the amount of fake money they ended up with. The resulting income distribution for one class is shown here:

wealth_dist_f2

 

In the other class, the rules were changed so that from each group, in subsequent rounds, the most successful student was given  2 additional cards, the second most was given 1 additional card, the least successful student had 2 cards taken away and the next least successful student had 1 card taken away. This was repeated at the end of round 2 as well. The resulting wealth distribution from this class is shown here:

wealth_dist_f1

In those class exercises we could expect that some students would be more talented poker players and that some students would be luckier than others, which did result in a less equal wealth distribution. What we (not so clearly from the graphs) observe is that if past luck can be used to influence future luck, the distribution of wealth become even less equal.

If luck significantly influences economic outcomes, what are the implications for economic analysis? Denrell, J., Fang, C., & Liu, C. (2014). Perspective—Chance explanations in the management sciences. Organization Science, 26(3), 923-940 argue that random chance can be the basis for explaining many “empirical regularities” (Denrell, Fang and Liu (2014) abstract).  So, what is the point of this post? Only that luck, in the form of random events, may very well motivate much of what we observe from markets, but that getting economics to admit this is not likely to happen very quickly.

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s