Memo to Jeff Bezos: Stack-Ranking is a Destructive Employee Practice

A survival of the fittest approach to hiring is a recipe for disaster.

Share with your friends

More share buttons
Share on Pinterest

By Alistair Thorpe and Rick O’Gorman 

Jeff Bezos, CEO and founder of Amazon, recently took some heat when the New York Times exposed working conditions and the corporate culture at his firm. ‘Ruthless’ and ‘demanding’ are two descriptors of the working environment, sink or swim. Amazon is not alone. Some of the leading recent startups have competitive employment requirements, a survival of the fittest approach. They want the best and push out the rest. It’s a simple notion to strengthening your company and the most efficient way to assemble optimally performing groups, organizations, and sports teams. Or at least that has been the dominant rhetoric behind models of group productivity within both the business and sporting industries. Stack-ranking and other business practices of individual selection have been widespread, from General Electric to Microsoft, and is a standard modus operandi in sports teams including the focus of this piece, the European soccer team, Real Madrid. However, the wisdom behind the application of these models, both in business and sport, is under scrutiny. To begin to see why, we turn to evolutionary biology.

In 1996, evolutionary biologist William Muir conducted a series of unusual experiments at Purdue University.1 Muir was looking to explore the various methods of group productivity with regards to egg production. He wanted to create a group of ‘Super-Chickens’ who would produce more eggs than any other coop. He followed the logic that many employers today tout: take the best individuals, put them in a group together, and then let the magic happen. Muir selected the most productive hens from each cage and bred the next generation from them. Muir also identified the cages that collectively were more productive at laying eggs in comparison to other cages. He then continued to selectively breed using these two separate groups and observed the levels of production.

The outcome of this study was striking; selecting the best group cages produced hens that thoroughly outperformed the line of individually more productive ‘Super-Chickens’. For the cage-selected line, after just five generations, the number of eggs per hen catapulted from 91 to 237, the mortality rate of the group crashed from 68% to 9%, and the hens also displayed improved wellbeing as a function of the reductions of pecking and negative social interactions.

Get Evonomics in your inbox

The Super-Chicken group did not fare so well. In fact, this line of hens had some other, rather less desirable qualities. They presented signs of aggression, violence, dysfunction and waste. There was an extremely high prevalence of fatal cannibalistic pecking within the group and general agonistic behaviors. Those in the cage who did not die from these cannibalistic attacks (there was an 89% mortality rate) were left with severe feather loss, life-threatening abrasions and other serious physical injuries. The hens were more intent on fighting amongst each other than doing anything productive! Hopefully that doesn’t sound like any workplaces you know…

So what happened? Why did the best egg-layers from the first generation yield something akin to the Gremlins of the eponymously named 80s movie? What Muir realized was that instead of identifying the most efficient hens, he had identified the hens that successfully conveyed the appearance of being the most productive. Those hens that individually produced the most did so by being adept at aggressively suppressing the other hens from laying eggs. Taking the more productive individuals meant taking the more aggressive hens. Breeding repeatedly from those which were most productive actually favored those which were most aggressive. Placing these hens together in cages led to extreme violence (only three of these psychotic hens actually survived!). Muir ended up running out of the Super-Chickens and had no choice but to end monitoring them and only continue with the other group. Ultimately, the process of selecting at the individual level took to an extreme the challenge of cooperation arising from individuals selected for selfishness.

The behaviour of the psychotic hens fits rather well with the normative assumptions of classical economic and game theory, which suggest that individuals will act selfishly in situations that afford them the opportunity.2 In group situations, individuals are consistently expected to identify, and act on, the dominant Nash strategy—the strategy that cannot be beaten. Just think of the classic ‘tragedy of the commons’, where people are predicted to free-ride on and exploit the contributions of others to a shared resource.3 Furthermore, the selfish actions of the Super-Chickens also support the theme of much evolutionary psychology from the 1960’s, which was based on the principle that individual interests will always outweigh the interests of the group.4Given an opportunity to benefit from the efforts of others, selection will favor those which seize the day.

However, the behaviour of the best group hens is not so easily explained by such classical theories. Nor are the endless examples of people who act selflessly and contribute to shared resources, thus prioritizing the group’s interests over their own (public television funding, charities, etc.). Intrigued by this conundrum, behavioural economists, Robyn Dawes and Richard Thaler, investigated why it is that people violate these principles of selfishness.5 Their primary conclusion was that, in general, the greater the incentive for group co-operation, the more likely the group is to work together. In addition to this, the recent re-examination of multilevel selection theory’s contribution to our understanding of group processes, after its wide rejection around the 1960’s,6 provides an explanation as to how in situations of between-group competition, altruistic groups can outperform selfish groups and lead to the favored selection of altruistic group members.7 Individuals do not rampantly exploit the commons; grazing commons as a shared community resource has often worked well. You just need the right incentives.8

This mismatch of incentives for individuals within a group is most certainly an area of concern for managers of all kinds. The relevance of these findings for management strategies in industry has gained some recent publicity, mainly following a popular TED talk by Margaret Heffernan which referenced Muir’s original experiments with specific focus on how traditional ‘pecking orders’ may not be the most productive organizational structure.

The Bezos, Ballmers and other top CEOs of the business industry are not alone in being troubled by the issues associated with building optimal teams. In team sports, the level of incentives for both the individual and the collective (team) reach unparalleled heights. Let us use soccer as an example. Teams are assembled with the sole purpose of challenging other teams for recognition and achievement. Yet within teams, players compete to be in the starting lineup, since squads are commonly comprised of 25 or more players but only 11 can play at a time. Furthermore, the most highly demanded players are able to hold sports clubs to ransom, with top players receiving over £300,000 ($434,385) a week in salary (e.g., Wayne Rooney, who plays for Manchester United in the English soccer top division).

As such, the modern world of sports provides researchers with a domain where they can explore which theoretical frameworks of group behaviour hold in practice. For instance, Swaab et al., established that while adding talented individuals to a group does initially result in improved performance, this relationship is non-linear and that, crucially, after a certain point the addition of too many talented individuals to a group has an inverse effect on group performance.9 Research on salary allocation shows one mechanism why this can occur, finding that pay inequality is linked to detrimental issues within teams, such that teams with highly unequal salary structures tend to also elicit more negative affect for their members, which can then lead to greater within-group problems.10 Furthermore, the Nobel Prize-winning economist James Buchanan – concerned with the aforementioned ‘mismatch of incentives’ within hockey teams – concluded that as the number of selfish players increases, so too does the likelihood for the other players to adopt the selfish behaviors.11 Similar to how the selfish behaviors of the Super-Chicken’s resulted in reduced productivity and negative intergroup dynamics, studies such as these have illuminated how, in sports too, critical numbers of selfish and talented individuals can negatively impact a team’s cohesiveness and output.

To observe whether the same line of reasoning is applied to assembling sports teams as was adopted by Muir in his chicken experiment, and that is also widely advocated within industry, soccer (commonly known as ‘football’ in many countries) offers a useful feature to explore the pursuit of super-chickens, its global player transfer system. In comparison to the heavily regulated draft systems in place in the NFL, NBA and MLB, soccer allows a transfer system of players from club to club (franchise) that operates more or less as a free market* and which affords comparisons with standard employment practices (due largely to the European Court of Justice’s 1995 “Bosman ruling” and the legal requirement for free movement of workers within the European Union). Transfers usually involve the trading of a player for cash compensation. The latest record was just established this summer: £89,000,000 (about $117,761,000/€104,750,428) for Paul Pogba to transfer from the Italian club, Juventus, to Manchester United in England.

Perhaps the best analogy of the Super-Chicken approach in football is the notorious ‘Galactico’ policy at Real Madrid, developed by Real Madrid’s president Florintino Perez in 2000, who sought to endow his club with the world’s best, and most famous, soccer players in order to achieve instant sporting and commercial success. To provide you with both a sense of the scale of the Galactico policy, and Real Madrid’s commitment to it, between 2000 and 2013 Real Madrid consecutively broke their own previously held world records for transfer fees spent on a player (See Table 1 below).

Table 1: The Galactico Policy: World record fees payed to the top players. Current exchange rates as of 15/09/2016 for EURO and USD.

The main issue for Real Madrid over this period has been that this aggressive policy of high profile recruitment has not exactly delivered the sporting success they would have expected. The Spanish league in which they compete is historically dominated by two teams: Real Madrid and Barcelona. Since Real Madrid embarked on this strategy they have seen a return of 18 major trophies for an expenditure of £924 million ($1.22 billion /€1.11 billion). Meanwhile for the same period, their main rivals, Barcelona have recorded 28 major trophies at a cost of £641 million ($847 million/€769 million). Having to watch their main rivals claim 10 more trophies at just over two thirds of the cost is clearly not what Perez had in mind with his Galactico policy. While many have opined on the differences between these two Spanish giants, a prevalent theme of much discussion on the two clubs is the recruitment policy. Barcelona are routinely praised for their unrelenting focus on developing young players from their academy, ‘La Masia’, and instilling in them a philosophy of teamwork, intelligence and commitment. The Director of La Masia has explained that, for Barcelona, when it comes to a star performer, “we try to give him focus without losing that spontaneity. He mustn’t lose that individuality, but he has to know he’s playing in a collective sport” (my emphasis). Barcelona’s club motto “Més que un club” – (More than a club) serves as a powerful statement of how their values and recruitment model differs from their Spanish rivals. In contrast, Real Madrid have had a high turnover of players, with many expensive signings being moved on to a new club, for a discount, a year or two later. Being a big signing in one year was no guarantee of avoiding an ignominious departure soon afterwards.

Upon review, the Galactico policy of acquiring (or selecting for) top players by paying extraordinarily large sums of money appears either likely to endow players with a ‘Super-Chicken’ label, or is likely to attract players with a pre-existing tendency to benefit individually at the expense of the team. Indeed, while the chance to sign for Real Madrid is a flattering opportunity for any player, the likely stack- ranking environment would seem more likely to attract certain traits. Accordingly, the relatively disappointing return of trophies and high turnover rate of world class players may be a consequence of the nature of the players recruited, or the behaviors which are coerced out of players by the high incentives to be the most productive or stand out performer. The high salaries, the media attention, and the risk of being sold to make way for the next galactico who is available on the market puts great pressure on players to perform in a way that justifies their fees and maintains their marketability. This could lead to a number of complex and counterproductive social dynamics within the group, much like the agonistic behaviors of Muir’s Super-Chickens. A further, and perhaps more indirect, outcome of such a policy is the external attribution of poor team performance. Certainly, this is the case at managerial level: club playing legend, Zinedine Zidane, is their 13th manager in as many years. In the same time, Barcelona have had five managers, and it might have been just three, only for ill-health to force turnover.

A concern for soccer should be that developments over the last twenty years have seen this policy become the rule rather than the exception. Where the model of exploiting soccer’s free market system to try and engineer teams comprised of superstar performers who are equally capable of achieving sporting success as well as generating revenue was once the exclusive privilege of one of world’s richest and elite soccer clubs, it is now the archetypal recruitment strategy for most sporting organisations. In fact, the policy looks to have spread to the extent that the Galactico term is now colloquially assigned to teams across a range of sports, such as cricket and rugby, which have adopted similar policies.

Attributed to the once technical director of Real Madrid, Arrigo Sacchi, is an insightful quote on this recruitment model “Today’s football [soccer] is about managing the characteristics of individuals…The individual has trumped the collective. But it’s a sign of weakness. It’s reactive, not proactive”. It seems that Sacchi saw in soccer the same thing that Muir discovered in his experiments 12 years earlier; teams constructed to function as a collective are the ones that will enhance the qualities of the individuals within it and prosper.

A shift away from rigid adherence to the assertions of classical theories is necessary both in business and sport whilst moving towards collective business practices and group level incentives has clear benefits to business organizations and sports teams. By reducing the incidences of intergroup conflict and competition, incentives targeted at group achievement and interdependence have a much greater potential for maximizing the output of individuals within groups.

Some companies have begun to pay heed. Recently, Microsoft abandoned its longstanding stack-ranking approach. It recognized that stack-ranking was undermining team cooperation, employees withheld information to avoid damaging their own rank, sought teams where they could rank better, and ensured new team members failed. As an outcome to stack-ranking, it seems obvious. Yet many companies pursued it, just as breeders choose super-chickens. For Jeff Bezos, he may not agree. Amazon is a hugely successful company, at least in sales and turnover. It is not a particularly profitable company. When Microsoft was the behemoth of its domain, its aggressive policies eventually led it into trouble, barely avoiding being broken apart. Enron thought it had the smartest guys in the room. It’s endemic corruption ultimately crumbled the company. A company led by super-chickens may not be the best long-term strategy.

In sport, a similar question can be posed. Assembling soccer’s equivalent of super-chickens offers no guarantee of success. Indeed, time and again teams assembling a ragtag group of A-listers have performed less than the sum of their parts. Teams built slowly, with care, and low turnover, often shine, at least until their better players are prised away. It takes time, patience, and belief. In both business and sport, short-term results weigh heavy, and the challenge is to convince investors and supporters to have patience. Maybe that is the real challenge.

*The most notable deviations being the restrictive conditions applied by governing bodies to limit the movement of players between employers to specified transfer periods, and that the transfer of a player under contract may not take place without the agreement of the player and both buying and selling clubs.


1. Craig, J., & Muir, W. (1996a). Group selection for adaptation to multiple-hen cages: beak-related mortality, feathering, and body weight responses. Poultry Science, 75, 294–302. doi: 10.3382/ps.0750294; Craig, J., & Muir, W. (1996b). Group selection for adaptation to multiple-hen cages: behavioral responses. Poultry Science, 75, 1145–1155. doi: 10.3382/ps.0751145

2. Baumol, William (1952). Welfare Economics and the Theory of the State. Cambridge, MA: Harvard University Press; von Neumann, J., & Morgenstern, O. (1947). Theory of Games and Economic Behavior. Princeton, NJ: Princeton Univ. 2nd ed.

3. Kerr, N. (1983). Motivation losses in small groups: A social dilemma analysis. Journal of Personality and Social Psychology, 45, 819-828; Kerr, N., & Brunn, S. (1983). Dispensability of member effort and group motivational losses: Free-rider effects. Journal of Personality and Social Psychology, 44, 78-94; O’Gorman, R., Henrich, J., & van Vugt, M. (2009). Constraining free-riding in public goods games: Designated solitary punishers can sustain human cooperation. Proceedings of the Royal Society B, 276, 323-329.

4. Williams G C. (1966). Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought. Princeton (NJ): Princeton University Press.

5. Dawes, R., & Thaler, R. (1988). Anomalies: Co-operation. Journal of Economic Perspectives, 2(3), 187-197.

6. Wilson, D, S., & Wilson, E, O. (2007). Rethinking the theoretical foundation of sociobiology. Quarterly Review of Biology 82: 327–348.

7. Wilson, D. S., & Wilson, E. O. (2008). Evolution “for the Good of the Group”: The process known as group selection was once accepted unthinkingly, then was widely discredited; it’s time for a more discriminating assessment. American Scientist, 96(5), 380-389; Wilson, D.S., van Vugt, M., & O’Gorman, R. (2008). Multilevel selection and major evolutionary transitions: Implications for psychological science. Current Directions in Psychological Science, 17, 6-9; Eldakar OT, Gallup AC. The Group-Level Consequences of Sexual Conflict in Multigroup Populations. de Polavieja GG, ed. PLoS ONE 2011, 6(10): e26451. doi: 10.1371/journal.pone.0026451; O’Gorman, R., Sheldon, K., & Wilson, D.S. (2008). For the good of the group? Exploring group-level evolutionary adaptations using Multilevel Selection Theory. Group Dynamics: Theory, Research, and Practice, 12, 17-26; Wilson, D. S. (2008). Social semantics: toward a genuine pluralism in the study of social behaviour. Journal of Evolutionary Biology, 21(1), 368-373.

8. Kniffin, K. M., & Wilson, D. S. (2005). Utilities of gossip across organizational levels. Human Nature, 16(3), 278–292.

9. Swaab, R. I., Schaerer, M., Anicich, E. M., Ronay, R., & Galinsky, A. D. (2014). The Too-Much-Talent Effect Team Interdependence Determines When More Talent Is Too Much or Not Enough. Psychological Science, 25(8), 1581-1591.

10. Kniffin, K. (2009). Evolutionary Perspectives on Salary Dispersion within Firms. Journal of Bioeconomics, 11: 23-42.

11. Buchanan, J. M. (2000). Group selection and team sports. Journal of Bioeconomics, 2(1), 1-7.

Donating = Changing Economics. And Changing the World.

Evonomics is free, it’s a labor of love, and it's an expense. We spend hundreds of hours and lots of dollars each month creating, curating, and promoting content that drives the next evolution of economics. If you're like us — if you think there’s a key leverage point here for making the world a better place — please consider donating. We’ll use your donation to deliver even more game-changing content, and to spread the word about that content to influential thinkers far and wide.

 $3 / month
 $7 / month
 $10 / month
 $25 / month

You can also become a one-time patron with a single donation in any amount.

If you liked this article, you'll also like these other Evonomics articles...


We welcome you to take part in the next evolution of economics. Sign up now to be kept in the loop!