Increasing Returns and Path Dependence in the Economy by W. Brian Arthur
The University of Michigan Press, 201 pages, $21.95
Review score: *** out of *****

And the Social Science, -- not a "gay science", but a rueful, -- which find the secret of this Universe in "supply and demand," and redues the duty of human governors to that of letting men alone, is also wonderful. Not a "gay science," I should say, like some we have heard of; no a dreary, desolate, and indeed quite abject and distressing one; what we might call, by way of eminence, the dismal science.

Thomas Carlyle, Occasional Discourse on the Negro Question, Fraser's Magazine, 1849

In 150 Years and Still Dismal!, David M. Levy points out that the above quote, which made the description of economics as "the dismal science" famous, is from a horrible screed defending racial slavery by Carlyle. Although the racist origin of the quote is commonly forgotten, the label stuck.

Like psychology, economics claims to be a science. Economics has managed to do something that has largely failed psychology (out side of neural network research): dressed up theory with vast amounts of mathematics.

Despite all the mathematics, economics has historically failed to fulfill the central goal of science: to describe and predict. The description of economics as "the dismal science" rings true for many people because the science of economics has largely failed to predict anything. Professional economists rarely, if ever, succeed in predicting economic expansion or recession. Even on critical policy issues like money supply, inflation, interest rates and government spending, there is no agreement among theorists.

The operations of the US and world economy is complex, so perhaps we can forgive the science of economics for its failure to provide predictive value. Markets are not as complex as the economies of large industrial countries and there is a vast economic literature describing markets. At least for some markets, classical economics does a good job of describing and predicting outcomes. For example, to take a page from Marx (Das Capital), let us consider the market for shoes. When initial innovation takes place, like the development of athletic shoes, large profits will accrue to innovators like Rebock or Nike. Over time other competitors will enter the market and athletic shoes will become more of a commodity. Prices will drop and profits will decrease. Success will go to the lowest cost competitor (which is why athletic shoes are produced in countries with low labor costs). Economists call this a decreasing returns market. Such a market is ruled by pure competition and tends toward an equilibrium.

I don't know too much about the prevailing theory about financial markets but, from what little I know, it continues to maintain the approach established by classical economics. This means that financial markets are envisaged as playing an essentially passive role; they discount the future and they do so with remarkable accuracy. There is some kind of magic involved and that is, of course, the magic of the marketplace where all the participants, taken together, are endowed with an intelligence far superior to that which could be attained by any particular individual. I think this interpretation of the way financial markets operate is severely distorted. That is why I have not bothered to familiarize myself with efficient market theory and modern portfolio theory, and that is why I take such a jaundiced view of derivative instruments which are based on what I consider a fundamentally flawed principle. Another reason is that I am rather poor in mathematics.

It may seem strange that a patently false theory should gain such widespread acceptance, except for one consideration; that is, that all our theories about social events are distorted in some way or another. And that is the starting point of my theory, the theory of reflexivity, which holds that our thinking is inherently biased. Thinking participants cannot act on the basis of knowledge. Knowledge presupposes facts which occur independently of the statements which refer to them; but being a participant implies that one's decisions influence the outcome. Therefore, the situation participants have to deal with does not consist of facts independently given but facts which will be shaped by the decision of the participants. There is an active relationship between thinking and reality, as well as the passive one which is the only one recognized by natural science and, by way of a false analogy, also by economic theory.

George Soros, The Theory of Reflexivity

Equilibrium theory has also been applied to stock and commodities markets. Here classical economic theory states that information in a market spreads rapidly. Further, markets are made up of rational traders who logically attempt to maximize their profits. For example, everyone who trades shares in Microsoft knows the current status of the government anti-trust suit. Classical economic theory states that Microsoft shares will always be properly priced based on current market information. Following this theory, a trader cannot beat the market and market prices follow a "random walk" (see A Random Walk Down Wall Street by Burton G. Malkiel).

Perfect market theory and economic equilibrium take a Newtonian view of markets. Markets can be described with tools like calculus, which describe smooth functions. As Brian Arthur points out in his talk The End of Certainty in Economics, classical economic theory dates back to the English and Scottish enlightenment of the mid-1700s. The greed to maximize profit in a market leads, in this Leibnizian view, toward efficient markets and a better outcome. The idea of market efficiency is particularly dear to some conservatives in the United States. In this view of the markets, the technologies that dominate markets are always the best (in this best of all possible worlds) because they are are efficiently selected in the marketplace. Following this line of reasoning, Microsoft is the dominate software company because they provide more innovation and serve their customers better than competing companies.

Although classical economics does a reasonable job at describing some markets, it does a poor job of describing many of the markets encountered in the modern world. As George Soros points out, any professional market trader knows that modern economic theory is incomplete. Markets are not efficient but go through booms, busts and fluctuate widely based on factors like the psychology of the market players.

In the last twenty five years or so a new view of economics has started to take hold. Instead of viewing economics through the lens of Newtonian mathematics, this new generation of economists views markets as dynamic, complex, evolving systems. These systems have non-linear dynamics with solutions that cannot be calculated without the aid of a computer. One of the pioneers in this area is W. Brian Arthur, who is currently the Citibank Professor at the Santa Fe Institute in Santa Fe, New Mexico.

Prof. Arthur's book Increasing Returns and Path Dependence in the Economy is a collection of Arthur's papers published in the 1980s and 1990s. As the title suggests, the papers cover two areas that are abhorrent to classical economics: increasing returns and path dependence.

Increasing Returns

Classical economic theory states that returns (e.g., profits) decrease. Profit in a market will be driven down by competition and the lowest cost producer will prevail (with thin margins). This is the classic view of commodity markets like shoes or dynamic RAM chips. As Arthur points out, it is also obvious that there are cases of increasing returns, where a producer gains market dominance and large profit margins over long periods of time.

Increasing returns (profits) are a product of a technological or knowledge advantage. For example, the design and manufacture of modern microprocessors requires a team of skilled logic designers and an even larger team skilled in the arcane art of semiconductor manufacture. Increasing returns are possible when a company like Intel is able to gain early market dominance. The lessons learned in the design and manufacture of the current microprocessor can be applied to the next product generation. The dynamic feedback of knowledge and capital allows a market leading semiconductor company to design more complex chips and build more advanced semiconductor foundries, making it difficult for a competitor to gain market share. This can lead to something approaching monopoly profits (increasing returns).

In the article reprinted as Chapter 5, Information Contagion, Brian Arthur and David A. Lane point out that increasing returns also exist for a market leader because prospective users are able to learn about the product. For example, the software used to design and simulate VLSI chips is expensive and complex. Prospective users commonly find out from existing users how their chip design projects have fared using the software before purchasing the product. Other products, with less market share, are at a disadvantage because it is more difficult for prospective buyers to find out about the product.

I returned, and saw under the sun, that the race is not to the swift, nor the battle to the strong, neither yet bread to the wise, nor yet riches to men of understanding, nor yet favor to men of skill; but time and chance happeneth to them all.

Ecclesiastes 9:11

I am hardly a biblical scholar. I owe this quote to the ultimate engine of serendipity, the Internet, and Andrew Odewahn article Babel Fish.

Path Dependence

Classical economic theory states that the best product dominates the market. The idea that the "better mouse trap" wins is also an idea that engineers love. Sadly for perfect market theorists and engineers founding start-up companies, the world does not operate by such simple rules. Random events, like chance meetings, the opportunity to hire the right people, or a good marketing campaign, can give a product an early lead in the market. The process of positive feedback and increasing returns can turn this early lead into market dominance. Inferior products can come to dominate the market even though superior alternatives exist.

Historical markets are full of cases where random events lead to market domination. Examples given by Arthur include the QWERTY keyboard and the VHS video cassette recorder's dominance over the Beta format. My favorite example of market dominance is Microsoft's MS-DOS (for Arthur's view on this see US Department of Justice Vs. Microsoft)

In 1980 IBM was out looking for an operating system for its coming PC. The legend is that [Gary] Kildall [the founder of Digital Research, the makers of MS-DOS] missed a meeting with IBM because he was out flying one of his planes. He could never live down that legend, but it wasn't entirely true. He was flying, yes, but he showed up only a little late. Then he talked all day and through the night on a flight with the IBM representatives back to their office in Florida. The sticking point: IBM wanted to pay a flat $200,000 license fee to get a royalty-free license in perpetuity. Kildall wanted more.

Bill Gates came up with a similar operating system. He gave DOS away to IBM for $50,000 and figured, correctly, that he could get rich by licensing the system to other computer manufacturers.

Gary Kildall, The DOS that wasn't by Jeffrey Young, Forbes.

At the time IBM dominated the computer industry and the IBM PC became the dominate desk top PC. IBM's choice of Microsoft's DOS established DOS as a standard and locked in increasing returns for Microsoft. The events that lead to this are random. DOS was certainly no better than Digital Research's CP/M. Digital Research later came out with DR-DOS, which was superior to MS-DOS in several areas. By this time it was too late and Microsoft was able to use what some believe was monopoly power to assure that DR-DOS would not gain market share (DR-DOS was purchased by Caldera, which settled an anti-trust suit with Microsoft for what was reported to be a nine figure settlement).

The early market adoption of DOS set up a positive feedback loop. Application developers targeted their software at DOS, which increased the dominance of DOS. Users became familiar with the DOS "standard", which further entrenched DOS. Yet as a product, DOS was beneath contempt. Both the user interface and the software API were terrible. Even in the early 1980s there were alternatives that were greatly superior to DOS. For example, UNIX System III was available and ran on systems that were broadly similar to the PCs of the era. Outside of the academic and scientific community there was little adoption of UNIX, since DOS had already been established as a standard.

Complex Systems

The view of Arthur and the new wave of economic theory, that markets and the economy are a complex system, means that there are few deterministic outcomes. The locations of cities and commercial centers (Chapter 6, Urban Systems and Historical Path Dependence) is based on random events. There may be several possible outcomes, which are impossible to predict in advance.

Any successful Silicon Valley CEO understands many of the points that Arthur makes. Arthur and his colleagues provide a formal foundation to describe modern market success. The equations developed in Increasing Returns and Path Dependence in the Economy allow computer models to be built and tested against reality.

The first step in science is to properly describe and model the world. The next step is the ability to predict outcomes based on these models. There may be good models, which act like real markets. But in a complex system it is unclear what their predictive value is.

Increasing Returns is no Guarantee

The power of increasing returns is no guarantee of market dominance. An inferior product cannot continue to dominate a market when there are alternatives that are much better. As desk top computers became more powerful, Microsoft would have lost their market dominance if they had not developed Windows 3.1 and Windows NT. The development of Windows 3.1 and Windows NT kept users locked into the Microsoft standard by providing adequate technology that was not so inferior to other alternatives that users were forced to switch.

For every story of a market leader using increasing returns to lock in market share there are other stories of companies that failed to keep market leadership. Nor does the power of increasing returns mean that inferior products always win. For example, in the 1980's Digital Equipment Corporation (DEC) became the second largest computer company in the US on the basis of its VAX mini-computer line. But DEC did not understand the power of Moore's law and the market impact of increasing microprocessor power. In the engineering community Sun workstations became a standard and Sun started to apply the power of increasing returns to display DEC. As Andy Grove noted, "Only the Paranoid Survive".

Market lock-In is also more complex than simple feedback. Adopting new technology is costly and users will not change unless they believe that the gain in doing so out weighs the cost of switching. For example, before the development of the Intel Pentium, Windows NT and Microsoft Visual C++ 5.0, Sun workstations provided the best platforms available for software development. This lead Sun to be the dominate platform for the EDA industry (VLSI design and simulation tools). Sun has fallen far behind and is now inferior to Microsoft. Compared to the professionally produced software shipped by Microsoft, Sun's software appears to have been developed by University graduate students. Yet Sun is widely used in the engineering community. Perfect market theory says that Sun should be abandoned by its users. But the cost of a Sun workstation is not that much greater than the cost of a similar system running Windows NT. The cost of adopting Windows NT out weighs the savings in many people's minds.

Why Only Three Stars

Like George Soros, I've never been that good at mathematics, at least not compared to people like Brian Arthur. I would have preferred a book with fewer equations and more discussion. As a result, I've given Increasing Returns three stars out of five. Actually, I believe that Brian Arthur might also have preferred a book with fewer equations. However, an economist cannot get papers published without mathematics (the more obscure and complex the better). Book reviews are subjective and this review is not ment to downplay Arthur's work. Nor does this review properly express my great admiration for Arthur's work. A mathematician or complex systems expert would justly rate this book higher.


Brian Arthur is by every account a very nice person who refuses to stoop to insult even when provoked. Never the less, he manages to stir up an amazing amount of controversy.

Arthur's theory of path dependence provides some of the intellectual underpinnings of the Justice Department's case against Microsoft. It should be path dependence and increasing returns is widely believed at a visceral level in the computer industry and at Internet companies. For example, companies like eBay, E*Trade and Amazon spend large amounts of money to establish a leading presence (e.g., they have attempted to set up a situation where they will achieve path dependent market dominance). Microsoft and their defenders have argued against path dependence. Against the evidence provided by MS-DOS, a product beneath contempt, Microsoft has argued that their dominance is based on providing the best product, not on, to use Arthur's terminology, path dependence and increasing returns.

The idea of path dependence and increasing returns argues that the market does not always yield the best of all possible worlds and that there might be a place for government intervention. Such a idea is anathema to some conservatives. One group that has published a letter arguing against path dependence is the The Independent Institute, which calls path dependence

"a fundamentally flawed economic theory that has no empirical evidence to support it and no evidence of consumer harm"

David J. Theroux, Founder and President, The Independent Institute, from a letter published in The Wall Street Journal, June 28, 2000, pg A27.

To counter what Oracle founder Larry Ellison saw as Microsoft's success in the court of public opinion in the Justice Department anti-trust suit (which Microsoft seems to be losing in Federal Court), Oracle hired an investigative firm, Investigative Group International (IGI), to investigate the level of funding provided by Microsoft to The Independent Institue and other groups. This involved going through the trash from these organizations (where was Faun Hall when she was needed?) As it turned out, Microsoft did fund a number of Microsoft supporters, including the Independent Institute, but Mr. Theroux claimed in the letter referenced above that this had nothing to do with their policy position. Obviously the success of a superior product like MS-DOS had nothing to do with path dependence and increasing returns. It was just another example of Microsoft "innovating for the customer".

Web References

Ian Kaplan - 6/2000
Revised: 10/2000

Book review table of contents

back to home page