A product of individual-level selection

The relevance of group selection to understanding behavior in groups—including in organizations like investment banks—is one of the most hotly-debated topics in human behavioral biology. (For example, see this recent debate on Edge.org). In essence, the debate is between individual selectionists who believe that behavior evolves primarily to benefit individual fitness (genetic survival and reproduction), and group selectionists who believe that behavior evolves to benefit individuals and/or groups. That may sound like a clear-cut debate, but it’s actually a very confused one. Unfortunately, a main reason why people argue so much about group selection is because there’s so much semantic and conceptual confusion, on both sides of the debate, about what group selection actually means.

To judge whether group selection is useful for understanding behavior in banks, then, we first have to define it clearly, or at least choose one of the three definitions of group selection that are commonly used: the good, the bad, and the cultural.

  1. The "good" definition: “A process of genetic selection that favors individuals who cooperate in groups, because cooperation allows these individuals to acquire fitness benefits (resources and status) that they couldn’t acquire acting alone”. This process can meet the minimal requirement to be considered group selection, if members of some groups have higher mean fitness than members of other groups. This kind of group selection is perfectly compatible with individual selectionism, because it says that individuals contribute to group efforts in order to benefit personally from group success. It also describes how people actually behave (see below).
  2. The "bad" definition: “A process of genetic selection whereby traits that disadvantage individuals in competition with other individuals in their own group, but that advantage groups in competition with other groups, can evolve.” Natural selection can operate simultaneously at the individual and group levels, and a self-sacrificial trait might benefit a group. If selection between groups were severe, and more successful groups were those which contained more self-sacrificial members, than a self-sacrificial trait could proliferate in the overall population, even as it decreased in frequency within each group. Although it’s a mathematical truism that a self-sacrificial trait could evolve in this way under the right selective circumstances, such traits are unlikely to characterize human nature, for two reasons. The first reason is theoretical: such a trait would, from the perspective of multi-level selection, be poorly designed (as I pointed out in my last post). It would cater to between-group selection pressures, but fail to utilize within-group pressures. It would therefore be vulnerable to being outcompeted by a trait that took advantage of both between-group and within-group pressures, for example, a trait which contributed to group goals while also avoiding being disadvantaged relative to lower-contributors in one’s group (“free riders”). The second reason is empirical: decades of laboratory and field studies on group behavior suggest that people are superbly designed not just to acquire the benefits of cooperation, but also to minimize their own costs of being exploited by free riders [1].
  3. The cultural definition: “A process of non-genetic selection whereby cultural traits, that advantage groups in competition with other groups, spread due to their advantageousness.” In other words, advantageous cultural traits (for instance, effective systems of criminal justice, national defense, or education) can increase in frequency because groups that display them tend to outcompete, or be imitated by, other groups. Such cultural group selection is compatible with individual selectionism: a group-advantageous trait of this sort is nothing more than what is commonly referred to as a “public good”, and individual selectionists accept that people have individually-adaptive, genetic adaptations for producing public goods. The cultural group selection angle on public goods just adds that some public goods tend to become more common in the world as the result of their being useful in intergroup competition.

So are any of these definitions of group selection consistent with the kind of behavior we see in investment banks? If we’re choosing between the two genetic kinds of group selection defined above, it’s obvious that the “good” definition is realistic, and the “bad” definition is fantastic. That is, bankers work for banks in order to make themselves money, not to sacrifice themselves for the good of the bank. (Yes, this point is obvious, but its very obviousness seems like a good illustration of what's unrealistic about the "bad" definition). What about the cultural definition of group selection —does that apply to banks? In theory it would, if banks were allowed to succeed or fail based on whether their corporate cultures encouraged sustainable organizational competitiveness, as opposed to encouraging employees to gamble away their firm’s future in the pursuit of short-term payoffs. In practice, however, government rescue missions have artificially prolonged the lives of banks whose cultures are dysfunctional and self-destructive in the long-term, and have thus interfered with the group selection of more sustainable bank cultures. 

(A version of this article will appear as the author's "Natural Law" column in the banking magazine Global Custodian).


  1. Price M. E. (2006). Monitoring, reputation and “greenbeard” reciprocity in a Shuar work team. Journal of Organizational Behavior 27: 201-219.

Copyright Michael E. Price 2012. All rights reserved. 

You are reading

From Darwin to Eternity

Did Religion Evolve to Benefit the Weak or the Powerful?

The paradox of who benefits most from religion

Surprised These Rockers Are Still Alive?

Why don’t we celebrate stars more while they’re still with us?

Star Wars, Christmas, and Constraints on Secular Influence

What's so special about religion (and Star Wars and Christmas)?