Economic cosmology - The invisible hand


Jason Collins


August 26, 2013

Adam Smith’s concept of the invisible hand is one of the more abused ideas in economics. Mentioned only once in The Wealth of Nations, and only then in the context of preferring domestic to foreign industry, the invisible hand has come to represent the idea that self-interest can improve the common good. The following phrase from The Wealth of Nations nicely captures the idea of the invisible hand (although it is not located with Smith’s use of the term):

It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own interest.

The invisible hand has taken on a life of its own since Smith’s nuanced book, with caricatured versions espoused and attacked along the political spectrum. In that context, the invisible hand is the second Western cosmology tackled by Gowdy and colleagues in their article Economic cosmology and the evolutionary challenge from the Journal of Economic Behavior & Organization special issue, Evolution as a General Theoretical Framework for Economics and Public Policy (I covered the first cosmology - the rational man - in an earlier post).

The first point they make is that the invisible hand operates in a context of human sociality and morals. Smith would not have disagreed with that claim, with any self-interested actions constrained by the norms of the group and the morals of the individual. But as for the first Western cosmology they considered - the rational man - Gowdy and colleagues suggest that selection at a higher level is required to explain these constraints.

In arguing this point, Gowdy and colleagues draw a distinction between social and non-social behaviours. They suggest that nonsocial behaviours that do not harm others can evolve through individual level selection. However, given that adaptations at a given level require selection at that level, social behaviours that increase the fitness of someone (or their genes) should not be expected to increase the fitness of the group. For that to occur, group adaptations such as group norms are required.

There are a few of points of interest this argument, although it is not clear to me how much these are due to framing rather than substance. First, consider two people who decide to trade with each other. Each has an object the other wants, they engage in the trade and both are better off as a result. Is this a social behaviour that has increased the fitness of the group? Using a multilevel selection framing, the two traders are a group, and their group is clearly better off. One of the traders may have gained an advantage over the other due to a larger gain from the trade, so within that group, one of the traders could be considered “weakly altruistic”. So, do we consider the trade a social behaviour and the group the two traders? And if so, can I also frame this as a simple self-interested action of each party to trade for something that they want more? That is of course a feature of the multilevel selection framework - the ability to frame it in inclusive fitness terms.

A second point is the distinction between fitness and economic outcomes. In an evolutionary sense, anything that increases your fitness decreases the fitness of others. Fitness is defined in relative terms. In an economic sense, actions can make everyone absolutely better off, as in the trading example above. So when Gowdy and colleagues state that social behaviours that increase the fitness of someone should not be expected to increase the fitness of the group, this does not necessarily rule out increasing the economic benefits obtained by the group. Of course, we could also measure economic benefits in relative terms, but that is not typically what modern advocates of the invisible hand concept mean when they talk of the group benefits of selfish actions.

From their argument that group selection is involved, Gowdy and colleagues seek to resuscitate the concept of the invisible hand by suggesting that the invisible hand is selection at the level of the group. Thousands of generations of group selection (both genetic and cultural) have shaped our psychological dispositions so that now there is no need to have any conception of the group in mind when we pursue our self-interest. Our group selection shaped self-interest tends to lead to group benefits, with these self-interested actions a very narrow subset of all the varieties of self-interest, most of which do not benefit the common good.

This is an interesting argument, although I’m not sure that I buy it. I agree that the set of actions that we can undertake to advance our own self-interest are constrained by other people, social norms and institutional frameworks. For example, in many societies, uncooperative behaviour can have severe costs. Further, we exhibit many constrained behaviours even when we are not actually constrained. But as I asked in my last post, if multilevel selection and inclusive fitness are just different ways of framing the same question, what does it actually mean to say that group selection was required? If it is simply a statement that if we use a multilevel selection framework and classify the groups in a particular way, most of the action will be at the group level, then it is a relatively controversial statement. But as I read the paper, I feel that the authors mean more than this, and are actually pointing to group selection in the older sense (see my last post on the rational man for some more discussion of this distinction). In other words, for the authors, this is more than just a question of framing._ _

The distinction between genetic and cultural group selection is also important. Cultural group selection is less vulnerable to critiques about the mixing in human populations that genetic group selection is subject to (although it still has plenty of critics), and these need to be addressed if they are implying older concepts of group selection.

Gowdy and colleagues close the article with the suggestion that the mix of self and other-regarding attitudes of humans, as shaped by individual and group selection, allow the division of labour and exchange to occur that drives economic activity. This regulation of competition and self-interest are the invisible hand that leads to the common good.

This is the practical implication of their article - the invisible hand requires constraint. Previously provided by group selection, constraint now needs to be provided by regulation. Gowdy and colleagues do not offer further detail on this point, but this is the ultimate purpose of this evolutionary foray into economics and where some interesting debates are going to occur.

My series of posts on the Journal of Economic Behavior & Organization special issue, Evolution as a General Theoretical Framework for Economics and Public Policy, are as follows:

  1. Social Darwinism is back - a post on one of the popular press articles that accompanied the special issue, a piece by David Sloan Wilson called A good social Darwinism.

  2. Four reasons why evolutionary theory might not add value to economics - a post on David Sloan Wilson and John Gowdy’s article Evolution as a general theoretical framework for economics and public policy

  3. Economic cosmology - The rational egotistical individual - a post on John Gowdy and colleagues’ article Economic cosmology and the evolutionary challenge 

  4. Economic cosmology - The invisible hand (this post) - a second post on Economic cosmology and the evolutionary challenge 

  5. Economic cosmology - Equilibrium - a third post on Economic cosmology and the evolutionary challenge

  6. Design principles for the efficacy of groups - a post on David Sloan Wilson, Elinor Ostrom and Michael E. Cox’s article Generalizing the core design principles for the efficacy of groups