Micromotives and Macrobehavior



ISBN: 978-0393329469

READ: 2016-09-22

AUTHOR: Thomas C. Schelling, Nobel Prize for Economic Studies in 2005



This book taught me to think a little more analytically about what happens in standard situations when individual, free choices lead to aggregate behaviours that nobody wants.
It is a classical problem of economics; Schelling warns the reader that every depicted function is nothing more than what it is – a model, an ideal scenario, free from its subtle, but no less important features. Schelling’s functions aim at “illustrating the kind of analysis that is needed, some of the phenomena to be anticipated, and some of the questions worth asking”; typically, they involve two-person scenarios as the notorious prisoner’s dilemma, or they focus on broader dynamics where the object of study is equally ‘simple’ – behavioural contrapositions between blacks and whites, rich and poor, young and old, etc.

Given that no underlying principle could be extracted from such a broad array of phenomena, Schelling lays out an interesting overview of macro-features: one is when phenomena occur in pairs as the previous dichotomies do, one when populations are guided by a principle of conservation (for example: no matter how hard you try, you can never get rid of the worst 10% employees, for the statistical feature of being ‘the worst’ is independent of the members of the system – the only solution will be to close your office), or move through a semi-closed system; Schelling then talks about complementary population sets as the two sexes or two ‘races’ (Schelling writes his papers in the Seventies, when racial issues were prominent in the US), and that “the independent variable in a system of behaviour often proves t be the sum of the dependent variables in a system”, and “people react to a totality of which they are part”.

Famous “critical mass” phenomena are just part of the family. One of the funny things about them, is that “even if one of the outcomes is unanimously chosen, we cannot infer that it is preferred from the fact that it is universally chosen.” To appreciate such an intuitive statement, graphics are an extremely useful tool, even though a little unwelcome for those who don’t like to see reality through straight lines.
S-shaped lines do help a lot, though. In “Thermostats, Lemons, and Other Families of Models” Schelling analyses how different scenarios may play out depending on which metrics we choose and act from to define phenomena. “If absolute numbers are what matter … the activity is likely to be self-sustaining in a large group but not in a small one.” Think of a typical dynamic, where people would go to that bar only if n. people are hanging out.
“If it is proportions that matter … there is the possibility of dividing or separating populations.” Speaking of language accent or fashion, for example, separating a population would have the effect of reshaping proportions, and make some behaviours more or less easily adopted.

An ethical and much actual note comes from the discussion of commons. Social contracts are sometimes blamed for not really being able to solve our problems. Schelling shows that such view is superficial, and that in the end having rules, being good or bad, is better than none.
In a classic problem of electrical overload, that half of the population which undergoes voluntary restriction may well be angry at the free riders. Nevertheless, even though free riders are better off than the other half, “the cooperative half may be better off for having found a way to make themselves cut back in unison.” Schelling here maybe suggests that ‘being better off’ may not be just to exert one’s individual choice unrestrainedly; there is a value, though less individual and thus less tangible and measurable, in the ability of acting ‘in unison’. And that ability, I’d further suggest, may well come in hand in the future.

Speaking of segregation and its dynamics, Schelling strongly opposes the view that such aggregate phenomenon has any social efficiency. Just as romance and marriage influence the aggregate genetic treasure that we’ll bring on, just as depression and inflation “do not reflect any universal desire for lower incomes or higher prices … The hearts and minds and motives and habits of millions of people who participate in a segregated society may or may not bear close correspondence with the massive results that collectively they can generate.” Now that’s a strike to social change efforts! 1)Other economists have taken this issue a little forward, and showed that the individual choice of giving up on one egg would decrease the total production by 0.91 eggs. More details could be found in Doing Good Better by William MacAskill and Expected Utility, Contributory Causation, and Vegetarianism by Gaverick Matheny.

As pointed out earlier, it is exceedingly hard to tell what’s behind individual decisions just by analysing aggregate phenomena. Certain configurations though may be mechanically produced by playing around with sets of preferences, as Schelling does in the classical simulation of neighbourhood segregation of pennies and dimes. From an initial, almost balanced disposition of dimes and pennies, assigning fairly human preferences as ‘staying with more than 3/4 of the same type, or leave’ turns the board to an equilibrium which is evidently clustered. Playing around with the coins will show that a distinction between integrative and separative ‘behaviours’ is almost impossible to draw.

Finally, a quite intriguing poke at randomness: imagine a binary division scenario with complementarity, one among the features laid down at the beginning – sex: people would be able to choose in advance the sex of one’s child. Most would likely prefer a 50-50 distribution, but individual choice may be driven by:
1) wanting a boy or a girl, while badly wanting the 50-50 population ratio;
2) wanting a child of the scarcer sex for some advantages;
3) wanting a child of the preponderant sex for equally conceivable advantages.
In such a scenario, not everybody will turn out to be happy. “The binary illustration is a vivid reminder that a good organizational remedy for severely nonoptimal individual choices is simply not to have the choice – to be victims (beneficiaries) of randomization – and thus to need no organization.”

References   [ + ]

1. Other economists have taken this issue a little forward, and showed that the individual choice of giving up on one egg would decrease the total production by 0.91 eggs. More details could be found in Doing Good Better by William MacAskill and Expected Utility, Contributory Causation, and Vegetarianism by Gaverick Matheny.