Building on our first post, we explore why nuance matters more in economics than it does in basic science, where you can map out the entire path of a football with just a few numbers 

Scientific problems necessarily involve simplification.  Simultaneously analyzing the interactions between billions of different molecules or millions of unique consumers is impossible.  Some nuance has to go.  But if you figure out a clever way to simplify your problem, you’ll only lose distracting noise while keeping the signal that you care about.  You’ll simplify without oversimplifying.

Toward the end of The Death and Life of Great American Cities, Jane Jacobs asked what kind of problem a city is, what kinds of simplifications we can make about urban areas.  She answered her apparently rhetorical question by saying that a city is a problem of “organized complexity,” a problem that cannot be simplified beyond a certain point, lest you convince yourself that housing projects like Cabrini-Green are a good idea rather than the abject failures they inevitably turn out to be.  You can say the same thing about the economy: it’s a problem of organized complexity that is nevertheless frequently oversimplified.

The economist’s predisposition to oversimplify arguably has its roots in successful simplifications of basic scientific problems.   Early scientists made extremely simplistic assumptions.  They assumed that gases in a container always spread out uniformly.  They completely aggregated all of the mass in flying objects into one representative point.  Although unrealistic, these models predicted reality accurately and precisely.  To understand why economists typically stumble when they make extreme simplifications, assuming uniformity or pretending that there is only one representative consumer in a market, it's helpful to explore why other scientists can get away with this.

Uniformity works by choosing a perspective where you are analyzing things that are roughly identical.  Consider a container of gas.  Although it’s impossible to know where each gas particle is, the randomness in the system ensures that—at least in a certain range of densities—the particles are spread out in a roughly uniform manner.  It's not exactly uniform, but if we assume that it is, calculating major properties of a gas boils down to adding up the effects of tiny, identical volumes of air.  Patterns of proportionality that we see in real world data—double the volume, halve the pressure—fall out of this simplification.

Randomness is the key here.  Although gases don't disperse uniformly in the real world, there aren't any systematic deviations from uniformity.  There is just noise that we don't care about, that we can simplify away without qualms.

Complete aggregation involves pretending that there is only one representative type of something.  This simplification works splendidly when used in Newtonian physics.  All of the atoms in a flying football have different force vectors acting on them; gravity is pulling harder on the atoms at the bottom of the football than the top.  But these differences balance out, allowing us to pretend that the entire mass of the football is located at one representative point, its center of mass.  Despite losing the nuance of the shape of the football, we can still make accurate predictions about its path.

The equalization of opposing forces is the key here.  We can pretend that the football is at one representative location because the differences that this assumes away perfectly balance out.  The differences between the real world and the simplified model don't affect where the football will go.

Extreme simplifications like uniformity and complete aggregation are rarely appropriate in economics because the modern economy is a place of interaction and fast changes, not randomness and balance.  Moviegoers choose films that their friends recommend to them; they don’t uniformly spread out among all of the options at the theater.   There are numerous owners of Walmart stock, and when one of them sells their stock, lowering the price, others can see this as a bad omen and sell their stock as well, continuing the pattern.  Differences feed off of each other instead of cancelling out.

Nuance matters in economics because the economy is a problem of organized complexity.  Money doesn’t spread out randomly.  Income differences don’t balance out.  The economy is a place of consequential interactions and feedback loops.  It isn’t analogous to a container of gas or a ball in flight.  It’s analogous to the complex problems today’s scientists solve with nuanced computational models, not the simple problems we solved centuries ago.


Photo via Flickr