Posted:June 29, 2006

‘Reasonably Good Strategies’ in the Face of Failure

It is the sheer complexity associated with many decisions that defies the orderly application of the rational calculations of economic theory.

— Paul Ormerod

Paul Ormerod’s Why Most Things Fail: Evolution, Extinction and Economics
begins with the observation that failure is everywhere but is not the subject of standard analysis by the economics profession.  Since the advent of the modern corporation, 22% of the top 100 companies at any given time drop from the elite rankings in the next decade, 10% of all companies fail each year (granted, mostly the newbies), and 50% of globally successful companies go extinct within the lifetime of a modern human.  Why?

Ormerod brings a journalist’s and economic historian’s perspective to the question.  He snorts at general equilibrium theory as being fundamentally at odds with how the real economy works.  Standard supply curves?  NO.  Demand curves?  NO.  Setting price at marginal costs?  NO.  Perfect competition?  NO.  Perfect information?  NO.  Eventually rising marginal costs with greater production?  RARELY.

Indeed, his attack on conventional economic thinking caused the reviewer in the Financial Times to rejoin:

For the foes he attacks are straw men. The simplifying assumptions of perfect competition and general equilibrium analysis are just that: no serious economist is unaware of their limitations. Economic models are not reality, but tools to help us think about reality. Only late in the day does Ormerod reluctantly admit the difference between claiming that economic agents act in a certain way for certain reasons, and arguing that for simplifying purposes, they act as if they do.

Yet the book peters out with a limp call for companies to "Innovate! Innovate!" This sounds like the guff of management textbooks. Actually it is worse. Other books about corporate failure that reach the same conclusion have an organisational theory behind them. Richard Foster and Sarah Kaplan, in Creative Destruction, suggest that companies fail because they cannot reallocate resources as rapidly as markets. Clayton Christensen, in The Innovator's Dilemma, says established companies cannot embrace discontinuous improvements. Ormerod has nothing to say about the internal innovation process: his analysis is a system analysis.

I’m not surprised that professional economists would find this book lacking.  Yet for most of us that have only taken Economics 101 or mostly get our economics from pop economists on news and business programs, Ormerod raises important questions and brings forth important ideas.  There is indeed much that the conventional economics community has gotten wrong or has been too simplistic about.  This book is a very good re-entry into economics if you have been away for a while.

Ormerod’s major thesis — and what caught my eye in getting the book in the first place — is that there are many parallels between biological evolution and the behavior and extinction of firms.  For this, Ormerod especially points to both systems as being characterized by much uncertainty and complexity.  No argument there.

In knocking down the pins of conventional textbook answers, Ormerod covers topics such as game theory, bounded rationality, probablity theory and power laws, invoking names and characters such as Von Neumann, Hayek, Stiglitz and others along the way.  His thesis also allows him to explore the question of exogenous and endogenous causes with some freshness.

Strangely, though, he is virtually silent on the real commonality between economic and biological systems — namely the central importance of information — and gets into a very strange and indefensible analysis of punctuated equilibria and extinctions for species spanning 550 million years in comparison to visual correlations that create an "immediate impression of similarity" with graphs of firm extinctions over the past 83 years.  As Ormerod states,

The relationship that describes the connection between the frequency and size of extinctions in biological species is the same as that which describes the extinctions of companies.  The timescales differ dramatically, but frequency and size are connected identically [my emphasis] in both.

Hogwash.

Ormerod gravely mixes samples, causality and power curves.  Indeed, similar power curves can be constructed for Web site popularity or term frequencies in the English language (among many others), which surely have nothing to do with "truths" about the extinctions of economic firms.  Also, given Ormerod’s thesis, he could have gotten some additional interesting mileage from Iansiti and Levien’s business ecosystem theory of The Keystone Advantage.

There is certainly insight that uncertainty makes it difficult to foolproof against the future and is one reason why things fail, also certainly compounded by complexity.  I also think Ormerod is able to make a pretty good case for how acquiring knowledge can help buffer against these factors. 

So, I give this book 5 stars for readability and provoking thought; indeed, it has gotten me to get re-acquainted with recent economics literature.  But I can only give it 1 to 2 stars for its logical side due to its mis-labored analogies and poor analysis.

BTW, for an intro to more serious discussions about the relationships of evolution and economics, see writers such as Nelson and Winter, Galor, Mokyr, or Ziman, among others, or the listings at IDEAS: Papers on Economics and Evolution from the Max Planck Institute.

Schema.org Markup

headline:
‘Reasonably Good Strategies’ in the Face of Failure

alternativeHeadline:

author:

image:

description:
It is the sheer complexity associated with many decisions that defies the orderly application of the rational calculations of economic theory. — Paul Ormerod Paul Ormerod’s Why Most Things Fail: Evolution, Extinction and Economics begins with the observation that failure is everywhere but is not the subject of standard analysis by the economics profession.  Since […]

articleBody:
see above

datePublished:

Leave a Reply

Your email address will not be published. Required fields are marked *