Damn facts

from Peter Radford

It seems appropriate to mention increasing returns today.  After all, this is the day on which several of our modern titans of industry are appearing before Congress to respond to the concerns raised by the gargantuan size that their respective businesses have grown to become.  The CEOs of Apple, Amazon, Google, and Facebook are all under the gun to defend the enormous clout that each wields in the modern marketplace.  Today is the culmination of a long investigation by Congress into the problems, or perceived problems, that these four companies represent.  Whatever the outcome the simple fact that the four are being clustered into one band of rogues taints them with an aura of indecency we normally associate with the pharmaceutical, banking, or tobacco industries.  That’s bad company to keep.

But back to increasing returns.

It’s one of those topics that economists like to tuck away and discuss out of the glare of public gaze. It represents a considerable challenge to the foundation of contemporary economics.  Economists love letting us know that they know about the potential various errors that might devastate the core of what they believe, but they equally love sweeping such anomalies under the rug so as not to have to re-invent their discipline.

And the big four tech companies being hauled before Congress today are each a pretty good example of the problem.  More exactly, they are all good examples of the existence of increasing returns as a fact in the real world.  Facts can be so annoying when you want to preserve your theory.  Economists have made a habit of collecting awkward facts and discussing them offline so that the central theory they love so dearly can be preserved from embarrassment.  Those brave souls who set off to examine the awkward facts get great praise and often serious support from their less brave colleagues.  They are, after all, inoculating the entire discipline against the charge that it ignores facts.  You can almost hear them say: No it doesn’t, just look at Professor so-and-so over there, they’re checking it out.  See?  Economics is really diverse and inclusive.  Under their breath, of course, you can hear them muttering that they hope Professor so-and-so doesn’t get so far that the Big Theory has to be torn down and replaced.

You can get a hint of this in some odd places.  For instance, Ken Arrow wrote about it in his Foreword to Brian Arthur’s 1994 collection of papers published under the title “Increasing Returns and Path Dependence in the Economy”.  Commenting on how the English School, particularly Ricardo and John Stuart Mill, dropped the concern Adam Smith had for increasing returns he says:

“Other analysts in different traditions, especially the French metahemtician and economist, A.A. Cournot (1838), saw clearly enough the incompatibility of increasing returns and perfect competition and developed theories of monopoly and oligopoly to explain the economics system implied by increasing returns.  But this tradition acts like an underground river, springing to the surface only every few decades.”

Arrow goes on to mention luminaries such as Chamberlin, Robinson, Young, and Kaldor all of whom have dared to bring increasing returns back to the main table only to be rebuffed as malcontents or misfits.

Indeed you get the feeling from Arrow that his own attitude is ambivalent.  He tells us that those aforementioned luminaries developed the ideas “though far from completely”.  Damning praise.  Perhaps Arrow, one of the modern greats and with a reputation that would have made him impervious to criticism might have helped complete the journey?  But no.  He spent his capital on the silly nonsense called general equilibrium, hoping, no doubt, that no anomalies popped up to reduce his effort to dust.

I must give Arrow his due, he confesses that the onerous requirements needed to make neoclassical general equilibrium work — let’s all admit that those requirements are other-worldly — look as if they undermine the entire project’s credibility.  So the wonderful edifice he helped articulate and bring into being achieves not much at all.  The excact same criticisms can be leveled against it that its defenders leveled against central planning in the infamous “socialist planning debates ” of the 1920s and 1930s.  Arrow is being honest when he observes that if even a single person anywhere in the economy can set a price (as opposed to being a passive price taker):

“the superiority of the market over centralized planning disappears.  Each individual agent is in effect using as much information as would be required by a central planner”

So general equilibrium is not a statement about the real economy, it is a statement about what economists would like to talk about — notably artificial markets.  I have often remarked, and not always in jest, that economists do not study economies, they study economics.  The two are very different.  Not least because economies often contain damn facts that poke holes into preferred theories.

A slightly different manifestation of this problem is the infamous Friedman position on the reality of assumptions in modeling.  There’s nothing wrong with making what appear to be extreme assumptions if the subsequent analysis produces some insight of value.  Friedman himself says that the goal of science is to make “valid and meaningful predictions about phenomenon not yet observed.”   This, naturally, begs the question concerning which phenomena he speaks of.  Most people, I would wager, look at economics as the discipline focused on making statements about markets, industries, firms, and whole economies.  It is not, according to people like Friedman, at all about the economic behavior of individuals.  This is odd because the entire effort following in the Friedman tradition has been to theorize about the collective phenomena only as aggregates of individuals.  This leads to the absurdity of Thatcher’s claim that there is no such thing as society.  She was making an ideological statement.  But it was based upon firm Friedmanite foundations.  The problem with this is that the corpus of statements about the collective phenomena are built on knowledge and assumptions that are, according to Friedman, not within the purview of economics.  If we are to believe this, then so-called “micro-foundations” are merely speculation exogenous to the core of economics and cherry picked in order to facilitate desired statements about aggregate behavior.  This conundrum exists only because those following in Friedman’s footsteps want to avoid being accountable for any prediction of individual human behavior based on that speculative basis.  Individual behavior is, remember, not the subject of economics, it is just an assumption that economists make in order to examine the bigger picture.  So what if the facts belie the validity of that speculation?  We can leave that to the behavioralists — another worthy subgroup working on the fringes and inoculating the mainstream against the charge of ignorance.

I have, as I usually do, digressed somewhat.  But the existence of things like increasing returns and realistic individual behavior lurk constantly outside the fences of economics and spring up, to use Arrow’s analogy, periodically to embarrass the mainstream.  They never appear to barge all the way in.  Somehow economics manages to push them away and consign them to specialist study as interesting sidelines.  They are treated as oddities and not the norm, so the conversation can continue much as before.

Which means, ultimately, that economics stagnates unwilling to expand its core range to encompass facts that are difficult to scrunch into its perfectly competitive general equilibrating world.  What then is positioned as advance is simply an ever more detailed discussion of the same stuff.  Decade after decade.  This is not a novel observation.  Leontief, hardly an non-person in the galaxy of economic stars, agrees:

“If the great 19th-century physicist James Clerk-Maxwell were to attend a current meeting of the American Physical Society, he might have serious difficulty in keeping track of what was going on.  In the filed of economics, on the other hand, his contemporary John Stuart Mill would easily take up the thread of the most advanced arguments among his 20th-century successors.  Physics, applying the method  of inductive reasoning from quantitatively observed events, has moved on to entirely new premises.  The science of economics, in contrast, remains largely a deductive system resting upon a static set of premises, most of which were familiar to Mill and some of which date back to Adam Smith’s Wealth of Nations.”

One last thing, speaking of stagnation: the current burgeoning “trade war” between the U.S. and China is often viewed through the lens of Ricardian analysis.  Comparative advantage encourages trade between nations.  So standard economics tells us that trade is a “good thing”.  But the current trade between the two is not actually trade between nations.  It is a complex of privately determined, shareholder driven, supply chain relationships.  Where is the comparative advantage?  Where is the nation?  Who benefits?  Simple little models built in the Ricardian tradition fail completely to get at the current facts.


Damn facts.