SECRETARIA DE ESTADO DE ECONOMÍA,
MINISTERIO DIRECCIÓN GENERAL DE POLÍTICA ECONÓMICA DE ECONOMÍA SUBDIRECCIÓN GENERAL DE ECONOMÍA INTERNACIONAL
CUADERNO DE DOCUMENTACION
Número 45
Alvaro Espina Vocal Asesor 6 de Mayo 2003
CUADERNO DE DOCUMENTACIÓN 06052003 45 ¿Están muertas la revolución de la información y la nueva economía? Brian Arthur dice que no
1. Brian Arthur: Vita………………………………. 3 2. Estamos condenados a vivir en dos economías..... 4 3. Is the Information Revolution Dead?…………… 7 4. Coming from Your Inner Self …………………... 23 5. His Economic Theories………………………….. 46 6. The Force of an Idea……………………………... 60 7. Brian Arthur: Selected Papers……. 65
7.1. "Competing Technologies, Increasing Returns and Lock-in by Historical Events," Economic Journal, 99, 106-131,1989. [18 páginas] 7.2. "Positive Feedbacks in the Economy," Scientific American, Feb. 1990. [12 páginas] 7.3. "Bounded Rationality and Inductive Behavior (the El Farol Problem), American Economic Review, 84,406-411, 1994. [11 páginas] 7.4. Preface to the book: Increasing Returns and Path Dependence in the Economy, Univ. of Michigan Press, Ann Arbor, 1994. [9 páginas] 7.5. "Complexity in Economic and Financial Markets," Complexity, 1, 20-25, 1995. [14 páginas] 7.6. "Increasing Returns and the New World of Business,"Harvard Business Review, July-Aug 1996. [10 páginas] 2
7.7. "Process and Emergence in the Economy," introduction to the book The Economy as an Evolving Complex System II, edited by Arthur, Durlauf, and Lane, Addison Wesley, Reading, Mass, 1997. [14 páginas] 7.8. "The Economy as an Evolving Complex System II.," W. Brian Arthur, Steven N. Durlauf, and David A. Lane, (Eds.), Proceedings Volume XXVII, Santa Fe Institute Studies in the Science of Complexity, Reading, MA: Addison-Wesley, 1997. Review by Gerald Silverberg, Maastricht. [6 páginas] 7.9. W. Brian Arthur, John H. Holland, Blake LeBaron, Richard Palmer, and Paul Tayler, “Asset Pricing under Endogenous Expectations in an Artificial Stock Market”, December 1996 preprint. Final version published as pages 15--44 in W. Brian Arthur, Steven N. Durlauf, and David A. Lane, The Economy as an Evolving Complex System II, Santa Fe Institute Studies in the Sciences of Complexity, Vol. XXVII, Addison-Wesley, 1997. [29 páginas] This paper develops the Santa Fe Artificial Stock Market Model. 7.10. "Complexity and the Economy," Science, 2 April 1999, 284, 107-109. [5 páginas] 7.11. "The End of Certainty in Economics," Talk delivered at the conference Einstein Meets Magritte, Free University of Brussels, 1994. Appeared in Einstein Meets Magritte, D. Aerts, J. Broekaert, E. Mathijs, eds. 1999, Kluwer Academic Publishers, Holland. Reprinted in The Biology of Business, J.H. Clippinger, ed., 1999, Jossey-Bass Publishers.[6 páginas] 7.12. "Cognition: The Black Box of Economics," The Complexity Vision and the Teaching of Economics, David Colander, ed., Edward Elgar Publishing, Northampton, Mass, 2000.[7 páginas] 7.13. "Myths and Realities of the High-Tech Economy," Talk given at Credit Suisse First Boston Thought Leader Forum, Sep 10, 2000. [5 páginas] 7.14. “Is the Information Revolution Dead? If history is a guide, it is not.”, Business 2.0, March 2002 Issue. [8 páginas]
2
3
Brian Arthur Vita
Citibank Professor, Santa Fe Institute
Schumpeter Prize in Economics 1990; Guggenheim Fellow, 1987-88; Fellow of the Econometric Society Dean and Virginia Morrison Professor of Population Studies and Economics, Stanford; Professor of Human Biology, Stanford, 1983-1996 Member, Board of Trustees, Santa Fe Institute. Director, Economics Research Program, 1988-90 and 1995-96, Santa Fe Institute Ph.D in Operations Research, Univ. Calif. Berkeley, 1973; MA in Mathematics, Univ. Mich., Ann Arbor,1969
Research Interests
• Increasing Returns. I spent much of the 1980s developing a theoretical framework for economic allocation under increasing returns, in particular studying the dynamics of lock-in to one of many possible equilibria under the influence of small, random events. (Several of my papers are collected in my 1994 book Increasing Returns and Path Dependence in the Economy. ) High technology operates under increasing returns, and to the degree modern economies are shifting toward high technology, the different economics of increasing returns alters the character of competition, business culture, and appropriate government policy in these economies
• Cognition and Economics. Standard economics typically assumes economic agents who face well-defined problems, and formulate their actions via some form of optimizing, deductive logic. In actuality many situations in the economy are indeterminate--a prime example is the El Farol problem. I am interested in such "pockets of indeterminacy," and in formulating economic theory for how human agents "cognize" problems, and how they operate under indeterminacy
• Technology and the Modern Economy. In the new high-tech era, the economy evolves as technology evolves. But how exactly does technology evolve? Is an economy that structures itself around high technologies different from a standard manufacturing economy? How does the information technology revolution play out in the modern economy? These are some of the questions I am currently thinking and writing about.
Two Books
The Economy as an Evolving Complex System II, edited with Steven Durlauf and David Lane, Addison-Wesley, Reading, Mass., Series in the Sciences of Complexity, 1997
Increasing Returns and Path Dependence in the Economy, Ann Arbor, University of Michigan Press, 1994 3
4
El ‘padre’ de la Nueva Economía, W. Brian Arthur: Estamos condenados a vivir en dos economías
Entrevista Realizada por Jorge Nascimento Rodrigues en Los Angeles. 7 Feb.01
No es una figura conocida en los medios de emprendedores europeos de la cuarta ola. Como se sabe, el megáfono de la ‘nueva economía’ ha estado en manos de Kevin Kelly, el célebre editor de la revista "Wired", primera en levantar este estandarte.
Kelly realizó con éxito la 'agit-prop' del concepto, cuyos ecos llegaron al Viejo Continente, pero el ideólogo es un reputado investigador del área de economía a quien descubrimos casi de incógnito en California.
Sus 'papers', publicados en revistas tan diferentes como "Scientific American", "The Economic Journal" y "Harvard Business Review", dieron a la realidad emergente la argumentación teórica que le faltaba, a pesar del riesgo que W. Brian Arthur, el personaje en cuestión, corría al desafiar abiertamente a la ortodoxia economista de finales de los 80.
Es una persona modesta. No reclama ser el ‘padre’ de la nueva economía, pero a él le debemos haber tenido la osadía de desafiar las teorías convencionales en los medios académicos sobre la dinámica del capitalismo, basadas en la ley de los rendimientos decrecientes de los factores y en la teoría del equilibrio.
Las dudas de Marshall
Los que estudiaron economía aprendieron que, a partir de cierto punto, un aumento en los factores no aumenta el resultado. Más de un lado no da más del otro. Las productividades marginal y media decrecen a partir de un punto de inflexión y los rendimientos pasan a ser decrecientes. Por otro lado, habría siempre un punto óptimo de equilibrio que significaría el uso más eficientes de los recursos disponibles en circunstancias dadas.
"Debemos estas verdades a Alfred Marshall, pero el célebre economista inglés, ya en 1890, en sus Principios de Economía Política, manifestaba algunas dudas sobre algunas anomalías, pero no siguió el hilo sus preguntas, pues la realidad no lo exigía", nos dijo W. Brian Arthur, 53 años, que vive hoy en día en Palo Alto y tiene una oficina ‘prestada’ en Xerox PARC.
La anomalía fue denominada como mecanismo de "feedback" positivo, o sea que, un resultado dado tiene un efecto de megáfono sobre los ‘imputs’, generando un círculo virtuoso. Más da más, y cada vez más. De allí nace una dinámica de rendimientos crecientes de los factores, al contrario de lo
4
5 que se observaba en el capitalismo industrial.
Lo que Brian hizo, desde fines de los 70, confiesa, fue seguir meticulosamente los viejos interrogantes de Marshall y lo que, al principio, parecían ‘anomalías’, fue la realidad emergente "donde la teoría convencional no se aplica más". Subraya que se sentía, ya entonces, bien acompañado, pues, antes que él, aún en los 40 y 50, el Nobel sueco Gunnar Myrdal y el keynesiano Nicholas Kaldor habían identificado tal mecanismo.
Las investigaciones más sistemáticas que hizo a lo largo de los años 80 produjeron varios ‘papers’ científicos que naturalmente sólo una elite leyó. Brian Arthur los compiló en un libro editado en 1994, con el título de Increasing Returns and Path Dependence in the Economy (ISBN 0472064967). El argumento de base era éste: la nueva economía, asentada en el factor crítico conocimiento, ya no opera en base a las viejas leyes.
El va a volcarse ahora nuevamente en la escritura; para el próximo año se espera su nuevo libro titulado, como no podía ser de otro modo, The New Economy.
"La gente que está desde hace muchos años en el hi-tech o en Silicon Valley, percibe perfectamente lo que digo. Lo saben intuitivamente. El propio Bill Gates toma partido de eso.
La Sun uso recientemente mis teorías en la estrategia de lanzamiento del ‘Java’, nos explica Brian, que fue profesor de economía en la Universidad de Stanford hasta 1996 y que ahora está en la dirección del Santa Fé Institute, en Nuevo México.
El azar histórico
El otro pilar derrumbado por Brian Arthur fue la sacrosanta teoría del equilibrio. La batalla no era fácil. Incluso una mente abierta a la "destrucción creativa" como Joseph Schumpeter consideraba que una hipótesis de múltiples equilibrios era, desde el punto de vista científico, una aberración. La consideraba, recuerda Brian, como "conduciendo a un caos fuera de cualquier control analítico".
Ahora, la posibilidad de varios equilibrios está en la realidad, aunque no quede bajo el control de los dogmas de los economistas. No se puede predecir un equilibrio óptimo. No, dice Brian Arthur: "Hay varias hipótesis planteadas y, en general, son acontecimientos fruto del azar histórico que seleccionan una solución determinada que, a veces, ni siquiera es la mejor desde el punto de vista tecnológico. Esa ventaja selectiva inicial permite el desarrollo de una red de partidarios y de dependientes que refuerzan positivamente el posicionamiento de liderazgo, del que el ejemplo más fuerte es la red liderada por Microsoft e Intel".
Sucedió lo mismo con el nacimiento de Silicon Valley, fruto de un acto casual, aparentemente insignificante, de un rector de ingeniería, visionario, que en los años 30 resolvió prestar de su bolsillo un poco más de 500 dólares a dos alumnos que terminaron creando HP en un garage.
Más recientemente lo imprevisto golpeó en 1993 la puerta de dos jóvenes de Illinois que obstinadamente querían hacer una interface amigable para la World Wide Web, en contra de la propia opinión del creador de la Web. Fue así que Marc Andreessen y Eric Bina crearon el primer 'browser', y no les pasaba por la cabeza provocar un nuevo furor económico.
5
6
"Hay aquí un paralelismo con la moderna teoría física de no linealidad. Pequeños cambios, en los que no reparamos a primera vista, pero que suceden en momentos críticos, generan nuevas situaciones inesperadas", prosigue Brian Arthur, que busca después en el filósofo Jacques Monod la imagen de un encuentro feliz entre el azar y lo necesario.
La dualidad actual
Lo que resulta de esto es que hoy conviven dos realidades económicas distintas. A veces, incluso dentro de la misma empresa, que tiene actividades operando en la vieja economía y otras en la nueva.
"Para mi, esta dualidad no es ningún monstruo de siete cabezas. La realidad está llena de ejemplos de eso", subraya nuestro interlocutor. Y prosigue: “Grosso modo, podemos decir que la ley de los rendimientos decrecientes vive en la parte tradicional de nuestra economía, y que la ley de los rendimientos crecientes es típica de las áreas basadas en el conocimiento. La economía de nuestros días se bifurcó en dos mundos interconectados – son dos mundos con lógicas económicas diferentes”.
Donde nos deja un consejo: “no mezcle ajos con ojos. Son dos economías diferentes en el estilo, en el comportamiento, en la cultura. Exigen técnicas de gestión diferentes, estrategias y códigos de reglamentación distintos. Es un error insistir en que lo que funciona en una funcionará en la otra”, concluye W. Brian Arthur para convencernos de que, en este caso, no hay otra solución que la bigamia.
6
7
Transcript: Is the Information Revolution Dead? W. Brian Arthur, Andy Grove, and Lawrence Lessig speak at Business 2.0 Live! event in San Jose.
By Business 2.0 Staff, April 2002 Issue
CARTER: The Commonwealth Club Silicon Valley and Business 2.0, brought to you from the Tech Museum of Innovation in downtown San Jose. My name is Dennise Carter, and I will be your chair for this evening. We would like to welcome the listeners of WBSU-FM 89.1 in Brockport, N.Y., one of the more than 225 stations joining us nationwide for America's longest-running radio program. Now it is my pleasure to introduce Ned Desmond, editor and president of Business 2.0, who'll be introducing our guest speakers, and he will be the moderator for this evening's discussion. Ned.
DESMOND: Thank you, Dennise.
DESMOND: Before we get started, I'd like to thank the Commonwealth Club Silicon Valley and the Tech Museum of Innovation for their generous assistance in making this evening possible. Our subject tonight: "Is the Information Revolution Dead?" Now regardless of what we decide, I can assure you that our panel is alive and kicking, engaged, and ready to weigh in on this topic. So let me introduce them for you. On my far side is Brian Arthur. Brian is an economist, and he's a pioneer of the theory of increasing returns. He's a professor at the Santa Fe Institute, and he's author of the essay that appeared in Business 2.0, the March issue, on the subject that is the inspiration for this evening. To his side is Andy Grove, the chairman of Intel and a co-founder of that remarkable corporation. And to my right, Larry Lessig, a professor at the Stanford Law School. Larry is an expert on constitutional law and legal matters relating to the Internet. His most recent book, The Future of Ideas, was published just last year. Gentlemen, welcome.
DESMOND: So let's get down to our discussion. I think we all agree that the late 1990s and the first part of this century saw a remarkable boom in the technology world. There was an astonishing crop of startup companies; there were dramatically overvalued stock prices, in particular for tech companies; there was ravenous corporate IT spending; and there was a vast industry of pundits, like me, who predicted the glories of the future of technology. And all of a sudden, just like that, in a matter of months it seems that all of that magic has gone away. In Santa Clara County, unemployment today is 7.5 percent, up from 1.7 percent just a year ago. Tech investors are a disgruntled group of people who've lost billions of dollars and don't have much faith in technology, at the moment. Even as the economy recovers in the United States, it seems that big tech companies just aren't keeping pace with that; they're still lagging behind. And skepticism abounds. Even Peter Drucker was quoted recently as saying -- in Business 2.0 -- "The information business as a business really isn't going anywhere." And yet, in his essay in Business 2.0, Brian Arthur argues that numerous examples from economic history suggest that information technology has yet to see its best innings. The future, Brian argues, is still ahead of us and not behind us. So I'd like to begin with
7
8
Brian and ask, how does economic history support this optimistic outlook?
ARTHUR: Well, there have been about four major revolutions -- technological revolutions -- before this one in information technology: The early Industrial Revolution, where factory and mill system was brought in; this one was in England. A so-called British revolution in railways in the early and mid-1800s. Another one that historians have identified -- steel and the coming of electrification of factories, a huge revolution in the late 1800s; this was in Germany and the U.S. And still another one the beginning of the 20th century until about 1950, '60, 1970 -- mass production and the coming of cars. If you look at those previous revolutions -- now, by the way, we're in what you might loosely call an information revolution -- if you look at those previous revolutions, you see that they go in recognizable phases. I wouldn't say we want to clock them with a stopwatch, but typically they work -- they start from a fairly primitive set of technologies. The amazing thing is that the technology works at all. Early in 1825 early railways that were introduced in Britain -- there were still carriages drawn on rails by horses. Ten years later there's a massive back and forth in technology, lots of inventors and promoters getting in the business -- a time of real turbulence. Then the press catches on around 1840, the late 1830s, and starts to talk about almost a new economy -- new prospects for the economy. It would be a very different economy. By the mid-1840s and the railway revolution in Britain, people had started to invest big-time. There were lots of promotions going on, little branch lines were being put in everywhere, scrip -- which is shares that are parceled up into tiny units -- they were being sold in alleyways. Everybody and his brother and his uncle was getting in on a massive investment spree. So there's a huge mania in 1845 and then a crash -- 1847 -- almost predictable. The Economist magazine was around at the time and had predictably forecast that there would be a crash and predictably said, "I told you so," afterward. Then, of course, you go into a phase where people are looking for scapegoats -- there's government committees and so on. But, interestingly, in that revolution and the others that I've looked at, it's not dead after the crash. The technology doesn't come to a halt. You go into a massive buildout of technology after that. That's when railways really started to take off in Britain, and 10 years later in the United States -- huge buildouts, still enormous investments, and that initial technology that had got everybody overexcited and then crashed everybody's hopes goes on to be a driver of the economy, brings in a huge amount of prosperity and growth. You enter a golden age. And then sooner or later, in this case around 1870, 1880, and a little bit later in the U.S., the technology gets tired; all the investments have pretty well taken place; foreign competition comes in -- if you have railroad money, you want to invest it in Argentina or Russia or somewhere else -- and the technology, then, is mature. What interests me is that in this period where there's huge excitement about a new economy, new things happening, a mania, and a crash, things don't stop there. That's when the interesting things start. And after that there isn't so much massive investment. You don't see huge manias after that; in fact, there's little bits of excitement. What you see is a very sober time, a time of buildout, a time of people saying, "Yes, let's get serious about this technology." Large companies start to take over, and it's a time of quiet buildup but huge buildup and always of prosperity. And this, I think, is the time that we're entering now with respect to the information revolution. We've had our mania, we've had our crash, and I believe what we're seeing is that the technology is starting to mature. The actual tech, the base technologies, are in place, and I'm expecting the next 5, 10, 15 years to be periods of buildup into a new period of prosperity and growth for the U.S. economy based upon information technology itself.
DESMOND: Thank you, Brian. Andy, did you ever feel you were running a railroad?
8
9
GROVE: Sometimes. Sometimes we ran off the track. You want me to comment on this?
DESMOND: I would like you to. Yes, please.
GROVE: As I was listening to Brian, a couple things occurred to me. One is the interrelationship between the different waves of technology, evolution, buildout -- that each had something to do with the next. Railroads made -- didn't start with steel; they started with wooden rails. They propelled the steel industry into big time. The steel industry, in turn, was responsible for the development of the automobile industry. So one industry's growth becomes a platform for another industry. And I think we can see some similarities in technology where semiconductors gave rise to the PC mass-produced computer world; mass-produced computer world gave rise to the Internet -- without PCs there would be no Internet; and hundreds of millions of connected PCs and the Internet is going to give rise to other types of applications that we're not sure of, just the same way as cars were not foreseen at the time the Bessemer furnace was invented. So that's one comment. The other one is a point of difference. The global economy in the last century was confined to England, some parts of Western Europe, and to the United States. Today the economy is really much more globally distributed -- information technology and the growth of information technology is much more worldwide, and the action, in fact, is taking place outside the United States as much or more than in the United States or Western Europe. That's a point of difference that actually favors the speediness of recovery.
DESMOND: I see. Well, Brian is our optimist; Andy adds a little caution to optimism. Larry, you're ...
GROVE: I didn't.
DESMOND: Oh, you didn't. I'm sorry. I misinterpreted.
GROVE: No, I think there's -- what I'm saying is that, first of all, this platform effect, that each wave builds on the other one, fuels the development of the -- that Brian was talking about. So I'm confirming that. And, secondly, the global afterburner that we have in the information technology in today's age wasn't present in his time so that's a plus.
DESMOND: That's going to amplify the effect.
GROVE: Right.
DESMOND: Larry, you're not noted for being an optimist.
LESSIG: No. Pessimism.
DESMOND: Would you buy in to what you've heard today?
LESSIG: Right. So pessimism is my brand. Let me stay on brand here. So I completely agree with Brian in his description of history, but the difference is there were no lawyers then.
LESSIG: And so I think the real question we should be asking here is, Is the information revolution 9
10 murdered? That's the real question.
LESSIG: And the reason we should ask that question is the significant difference that we're facing now is that using the law and the power that intellectual property of a particular kind is giving concentrated industries -- the future development that you, I think optimistically, point to can be vetoed by last generation's -- by the dinosaurs of the last generation's technology. So the way the 19th and 20th century thought about distributing content gathers a bunch of intellectual property and says, "Ways that are different from our way, and ways that are different, and ways that undercut our monopoly power, we're going to stop." And the difference now is the future technologies that were allowed to rage after the railroad revolution and the car revolution are now illegal. These are illegal technologies we're talking about. And they're talking about making machines that Andy's company builds that don't have copyright police built into them -- illegal technologies. So I think the real question we've got to focus on now is not the great optimism of the pattern of history but particular powerful forces that can intervene and undercut that revival, which, I agree, we should expect but can be destroyed if these trends continue.
DESMOND: That's a very rich subject. It leads me right to my next question, which is what do we need to see happen to usher in the golden era that Brian spoke about and Andy feels is going to be amplified.
GROVE: Shakespeare provided a prescription for that.
DESMOND: And it is ...
GROVE: First kill all the lawyers.
DESMOND: We can just end it right there. It's been nice knowing you.
ARTHUR: Does that include law professors?
DESMOND: I don't know. What would you say, Larry?
LESSIG: Oh, it's such a guilty conscience I have here.
DESMOND: Well, clearly, you're a reformist. You're trying to save the profession from itself.
LESSIG: OK, so now I'm going to defend the lawyers.
LESSIG: The law in this area has traditionally been extremely balanced. We lawyers use the word "intellectual property," but we're licensed to use that word. We're trained that intellectual property is a balance between monopolies granted by the state to create an incentive to produce and access, and fair use guaranteed by the balance the law creates to ensure that follow-on innovators and creators can take advantage of the innovations that have been produced. That's the balance the law has traditionally struck. The problem today is that this word "intellectual property" has become captured by people like my friend Jack Valenti, who goes around talking about intellectual property not as a balance but as an extreme; not as something that we're supposed to be constantly restriking as technologies change to make sure it doesn't stifle innovation, but as a tool that the dinosaurs can use 10
11 to make sure there are no mammals in the future. So he talks about intellectual property as just like every other kind of property, and real lawyers listen to that and say it kind of rings hollow because it's not like every kind of property. There's nothing that says you have a fair use right to my car, and I guarantee it -- don't touch my car. There's nothing that says that my car will be turned over to the public after a "limited time." But the Constitution guarantees that there's such a thing as fair use to copyrighted materials and guarantees that intellectual property gets turned over to the public after a limited time, and that's not because the framers were communists, it's because they understood that intellectual property is different. And we've lost that sense of difference, and we've allowed very powerful industries therefore, I think, to veto the future that Intel and companies like that would produce.
DESMOND: Brian, in your research, have you encountered a similar sort of roadblock to progress in the course of other technological revolutions?
ARTHUR: Oh, absolutely. At the moment, it appears that Hollywood is in a raging battle with Silicon Valley, and I agree that the dinosaurs are trying to stop the arrival of mammals. Around about 1865, the coach makers and the coaching industry in England passed a law -- the famous Red Flag law -- and it said anybody with an automobile -- in those days they were steam automobiles -- had to, number one, obey a speed limit of 4 miles per hour ...
ARTHUR: ... and you had to have a man walking in front waving a red flag.
ARTHUR: Now, I mean, I have this grand vision of what freeways would look like if we adhered to that law. It did last on the books, though, about 13 years and was gradually modified. I do think that Larry's quite right that you can't let the new revolution fall into the hands of the people who ran the last revolution. That's a legal disaster and it's an economic disaster. You simply can't let the information technology revolution be hijacked or taken over by people from a previous regime. Why? Their interests are at stake, and they will do everything they can to have guys waving red flags in front of information technology.
GROVE: Is there an example that you can think of in history where the progress of technology was more than temporarily disrupted by previous-era representatives blocking it?
ARTHUR: I think every -- my experience is that every good idea always has its skeptics. In this case it's more than skepticism, though. These are powerful economic interests trying to act, quite understandably, in their own interest and, thereby, getting in the way. What strikes me, as an economist, is that every revolution you see in history needs some sort of infrastructure. You need a railroad system to run trains on. You need an interstate highway system to run automobiles and the whole mass-product economy on. And it takes a lot of time and an awful lot of political infighting for this to take place. The first proposals for the interstate highway system, to my surprise, were in 1924. Roosevelt was a big backer in 1938. And it wasn't until Eisenhower came along in 1954, 1956, that the Highway Act finally got passed. I think we're seeing a repeat of this. The huge fight here is over broadband. Broadband is going to be the infrastructure of the Internet and this digital economy we're talking about. Some people say, "Let's get on with the job" -- that's Intel, Silicon Valley. I thoroughly favor that, as an economist. Other people are saying, "Well, hang on. If we put in broadband, I might get ripped off. My industry stands to lose. Let's set up a lot of roadblocks." So it's -- I agree with Larry but it's not just a bunch of lawyers fighting. There's major, major stakes in 11
12 this one.
DESMOND: Um-hm. Let me ask this question. If you had to assign some weight to the different factors that are slowing down the forward march of technology today, would you say it's principally the lack of dissemination of broadband? Is it the resistance of the dinosaur industries? Is it the lack of real -- what Brian sometimes calls the amenity of technology? I mean, where do these different forces come into play, and how important are they as breaks on our forward motion?
GROVE: I'll take a crack at that. If you think of it at this moment in time, the impediment is the lack of availability of broadband. If you somehow magically made broadband available to everybody who -- let's take in the U.S. economy, in the United States -- everybody who wants it, you would instantly go from maybe something between 6 million and 10 million households being connected to the Internet by a better-than-dialup modem -- we call that broadband, in a bit of a misnomer -- we would jump up to maybe 15 million. After 15 million, things would come to probably a halt because people would be looking for motivation. The next 10 million people would be looking -- "What do I need this for?" or "Why do I want to pay $49 instead of $20? There's nothing on the Internet," so to say. Then we're going to switch over to the content availability. So at the moment there is a pent-up demand for -- that could double the broadband penetration in the U.S. As soon as it doubled, it -- and I'm kind of figuratively thinking of doubling it -- the content availability would come in as a limitation. So they are sequential.
LESSIG: I agree with that, and I would add to it a particular twist that the United States courts have intervened with recently, in contrast to what happened in the past. Historically we've seen many times where new technology has affected the delivery of content and the ability for people to gather revenues. The most important first example is with sheet music and the player piano. Sheet musicians made their money by selling copies of sheet music. Player-piano producers came along, they bought one copy, and then they Napsterized the sheet-music business by then selling many, many copies of piano rolls. The sheet musicians went to court; they said, "We're being Napsterized. Stop them from stealing our stuff." The Supreme Court said, "I'm sorry. The last doesn't apply." So it took Congress to intervene years later to strike a new balance. The same thing happened with cable television, where cable television -- a new technology for gathering broadcast and selling it to their customers. That's the Napsterization of broadcasting. Twice the Supreme Court was asked to shut this down over a period of 15 years. Twice the Supreme Court said the law doesn't apply. It took Congress 20 years later to sit down and restrike the balance. And then in 1976, Universal and Disney launched a lawsuit against Sony for the then current Napsterization technology, which was the VCR. Eight years later the Supreme Court said, "I'm sorry, your content's being Napsterized, but tough. When there's a major technological change, it's not our job to restrike the balance between content protection and access. That's a job for Congress. And so as long as there is at least a substantial, legitimate use, we're going to allow that new technology." The problem now is that courts don't behave like that. The problem now is you have a new technology. You take the replay TV technology -- which is a much better VCR-like technology for recording television shows -- that technology has got to go into a federal court and prove that it's a viable technology that, on balance, will not harm the copyright interests of the broadcasters. They tried to invoke a Napster -- the Napster defense, that would be a real loser ...
LESSIG: They tried to invoke a Betamax defense, saying, "Here's a couple legitimate uses, that should be enough under the Betamax case." But the court said, "No, we're going to require you to 12
13 prove that this technology, on balance, isn't going to do harm." Now that's like saying they're going to have a bunch of economic experts -- I'm sure they'll be very good, Brian -- but, still they'll be experts hypothecating about what the future is. It's worse than lawyers. And then you'll have a federal judge sitting down and deciding whether the technology would be allowed. Like a Soviet planner, but with better lighting ...
LESSIG: Now this is our innovation policy. Federal courts deciding which innovations will be allowed. And so when you say the next kick is going to be content -- the real problem is the venture capitalists know they're not going to fund any new venture capital project that has to do with the delivery of content unless the dinosaurs approve of it because they'll buy a lawsuit. And, of course, that's the last thing any venture capitalist wants to invest in.
GROVE: You're discounting the possibility of a consumer rebellion.
LESSIG: True. Right. And the last time we saw such a rebellion was ...
DESMOND: Napster.
LESSIG: Right. Where was the rebellion with Napster? I mean, 70 million users were all of a sudden told they were criminals, and what did they do? They said, "OK, we'll switch channels and watch ...
GROVE: Switched to Morpheus.
DESMOND: Switched to Morpheus, that's right.
LESSIG: Right. Well, that wasn't the rebellion that was going to get -- I mean, I hope for it, right. And maybe we can get it going to, say, rationalization here, but I haven't seen it.
DESMOND: We've heard about broadband, and we've heard about content. Have we covered all those bases, Brian, all the impediments?
ARTHUR: No, I think there's another one. I'm sitting here wondering if Larry wants to kill all the economists as well as ...
ARTHUR: ... if Andy wants to kill all the lawyers. There's another one. I thoroughly agree that those are major roadblocks to this buildout. But there is another one and that is that, put very crudely, the Internet as a technology doesn't work. It doesn't work very well. I'm trying to hook up from home yesterday on a standard telephone line -- I'm somewhat of a Luddite or a technology primitive -- and I can't do it. I can't get into my server easily. I get onto the server this morning, it doesn't work very well. The point is that at this stage of these revolutions, the base technologies are in place, but what it takes to make those technologies usable are a thousand and one small what I call amenity technologies. The arrangements that make them all workable. As Andy was saying when we had early trains, say around 1850, they worked on wooden rails covered with iron plate. It wasn't until you get steel railroads around 1870, Pullman cars, airbrakes, switches, signals, telegraph systems from one railroad¿s depot to another one that makes it all work. So the technology needs to meet us as consumers and work. The other thing is that we, as businesses, are users and all these 13
14 revolutions have to completely upend the way we do business. And completely re-create our businesses. In the 1890s, 1900, dynamos were coming along -- electric motors -- and it looked very easy. You could just put those into the factories that were powered by these enormous steam piston engines with pulleys and wheels and levers and all kinds of gizmos. But it turned out that to make that technology work, to make electrification work, you had to completely redesign factories. Now I think we're in the same position. Not only must the Internet work for us so that I'm not, as a consumer, just standing there, sitting there totally frustrated -- will it work, won't it work, can I get into my bank account, will this take me 10 minutes or can I just get in at the press of a button -- so that the whole technology can be taken for granted and become invisible, as John Seely Brown puts it. But also we have to -- to make this revolution usable, of businesses really speaking to each other and parts of businesses interconnected with each other, is going to require a very different type of business organization and we haven't a clue yet what that's going to be. So all of that's going to take time. It'll take several years. Maybe 5 years, maybe 10, maybe 15. And this is what really draws things out. But as that comes along, there will be, indeed, a massive buildout.
DESMOND: Andy, do you have some comment?
GROVE: Well, first of all, I took the time that Professor Arthur was talking to answer the pop quiz that I got from Professor Lessig.
GROVE: There was a consumer rebellion ...
DESMOND: But that's cheating.
GROVE: ... in the mid-'70s against Detroit and propelled the foreign car penetration to the United States from the 20 percent to the 50 percent range because people were dissatisfied with the car design that prevailed in the '70s -- the size of the cars, the gas mileage of the cars, the quality of the cars -- most importantly -- and caused a wholesale adjustment of the most entrenched, most traditional business environment, which is the car manufacturers, in the United States, to adjust to that...
LESSIG: OK, but when the consumers rebelled, they weren't called terrorists. When they rebelled, they weren't called thieves and criminals. When they said, "We want a different ..."
GROVE: They were by Lee Iacocca.
LESSIG: Well, by Lee Iacocca.
DESMOND: By Lee Iacocca, right.
LESSIG: But Lee Iacocca didn't have Jack Valenti backing him up. And the point is Valenti et al. have succeeded in painting this debate as a debate between property-loving Americans and thieves with ponytails, typically, right. Thieves who are out there just trying to undermine the American system through these new systems for distributing content.
GROVE: To which the ponytail, 20-somethings have stopped buying CDs, and for the first time in 20 years, the dollar revenue of the music industry has taken a step down to the tune of 5 percent a 14
15 year.
LESSIG: Right. Which backs up their argument in Washington to say, "We've got to stop the piracy and thievery," and allow Sen. Hollings to hold a committee where he completely, inappropriately, and, I thought, unfairly brutalized your colleague, Les Vadez, who had just asked questions about shouldn't we be thinking about consumer rights here and shouldn't we be letting the technologists work the problems of piracy out without government ban, regulation. The problem is now most Americans think this is about thievery, and we have failed -- those of us who believe in innovation and creativity have failed in framing this not as a battle between those who want to steal and those who want property but about whether we allow the future innovation to take off.
GROVE: Larry, you say most Americans think -- most Americans that matter in this question are the people who make up 80 percent of the music-buying public and I bet they don't think of themselves as thieves.
LESSIG: And they also don't vote. I mean, the fact is there's one, maybe two, maybe three...
GROVE: They vote with their credit cards.
LESSIG: They vote with their credit cards, but the fact is those who are passing laws are not Visa and MasterCard. Those who are passing laws are congressmen who are responding to completely apathetic ponytail people and people buying music who say, "I'm not going to waste my time with politics," and why should they? It's a complete waste of their life. And they're not responding to those people; they're responding to the people who are spending lots of money lobbying Washington to back up their dinosaur industries.
ARTHUR: What I'm hearing from my two colleagues here is if you have a ponytail call your congressman right now.
DESMOND: On that note, let me just make a 15-second interruption for the benefit of our radio listeners. This program is brought to us by the Commonwealth Club Silicon Valley and Business 2.0. The program's title: "Is the Information Revolution Dead?" Our guests are Brian Arthur from the Santa Fe Institute, Andy Grove from Intel, Larry Lessig from the Stanford Law School. My name is Ned Desmond. And let me take that opportunity to wedge a question in here because I can't resist and I told Andy I'd ask him this question. In hearings before the Senate commerce committee, Michael Eisner of Disney said that Intel, Dell, and Apple were basing their future on piracy of goods created by companies like Disney, and he went on to say that he was sick and tired of being finessed by Silicon Valley companies who promised cooperation and didn't deliver. So, Andy, are you a pirate and have you been finessing Michael Eisner?
ARTHUR: Please take off your eye patch before you answer that.
GROVE: Mr. Eisner followed the Jack Valenti school of histrionics in front of the committee. Intel believes in intellectual property; Intel is built on intellectual property; and Intel respects other people's intellectual property, as do I. And as to the question of do we finesse them or do we work with them -- we work with Disney as well as the other entertainment companies for six years trying
15
16 to forge a consensus on content product technologies and, so far, we have not succeeded.
DESMOND: What do they want from you?
GROVE: Do you know the story about bringing a rock? It's the kind of thing that I tell you, "Ned, bring me a rock." You look a little puzzled, turn around, go to the riverbed, bring me a rock, and I say, "No, not that rock, another rock."
GROVE: Some parts of the entertainment industry are playing "bring me a rock" with us.
DESMOND: Brian?
ARTHUR: Well, we've been making a little bit light of this but I believe, as an economist, I believe this is an extremely serious issue. To an outsider it may appear to be inside baseball or something of a tempest in a teapot where Silicon Valley's pitted against Hollywood, the content providers against the technology providers. But, actually, it is a very serious business because I believe that high technology and the information revolution as exemplified, say, by Intel and other high-tech companies, this technology and this revolution -- is beginning to drive growth and productivity in the entire U.S. economy. This is what's keeping the U.S. economy internationally competitive and keeping the edge of the U.S. economy sharp. If you put up obstacles and roadblocks to innovation in this area as the U.S. government, you do so at your peril. Other countries can take over, and there's plenty of historical precedent for that. In the 1880s, Britain lost its edge and America and Germany started to take over. There was still plenty of prosperity to come in Britain, but they ceased to be leaders. And I worry that if this sort of issue isn't settled in favor of innovation and in favor of buildout and not looking to the past to say what worked in the past and looking to the future, rather, to say what do we need to make all of this work for everyone in the United States, I worry if that's not settled that high tech will gradually lose its steam and just simply go abroad. Manufacturing went abroad in the 1970s and '80s with the rise of Japan and Korea. There's nothing magical about the United States, in my opinion, except its ability to innovate and innovate freely, and that's what this country's all about.
DESMOND: So if Jack Valenti were here, he'd say every day 350,000 films are downloaded illegally over the Internet, robbing Hollywood of its revenue and its profits. What does Hollywood need? What should Hollywood be satisfied with in the current environment?
GROVE: Hollywood needs to offer a legal and suitable, legitimate opportunity for downloading movies so the 350,000 people who go through all kinds of contortions that do put your little troubles of getting on the Internet trivialize -- make it trivial in comparison.
GROVE: But they do it because that's their only alternative. What is the legal alternative to doing it illegally? What is the legal alternative of downloading music more than a trivial percentage of the content available? What is the service that they have licensed to provide that legal alternative with? It's not available. So I would say to Mr. Valenti, "You're leaving me no choice. I either curl up and go to Tower Records and buy the $16 CD so I can have the one song that I want, or you drive me to Morpheus or the brothers of Morpheus.
DESMOND: So it's the question of providing a straightforward option to use the Internet to 16
17 distribute digital entertainment.
GROVE: Right.
DESMOND: Larry, do you have any thoughts on that?
LESSIG: I think we don't really know what the future distribution of content is going to look like in a market where there's going to be intellectual property properly respected and, I, too, agree; I believe in intellectual property but where it's properly respected. We don't know what that's like yet, and we ought to understand we're in a period of transition. And in a period of transition, as Brian was saying, the important thing is just to make sure we move through the transition quickly and easily. Now when we get to the end of this transition, I think we're going to have very different models as John Seely Brown also says, there's -- the different architecture of revenue for these companies will make it look very different from the way it looks right now for them to make money and they will make money in this new business. The diff -- the key reason why it won't be a happy march to this future where everybody's making money once again is that the same people won't be making money. Look in the music industry -- the highly concentrated industry -- five companies control over 80 percent of the distribution of music in the world. One might call that not a healthy industry. One might think it's just too concentrated, and one might say competition that the Internet could provide here would actually improve both quality and diversity of music and also the position of artists. Now if that's what the future delivers, that's going to be social gain -- social welfare gain.
GROVE: Larry, before the personal computer, five companies provided 80 percent of the world's computers.
DESMOND: Before the PC?
GROVE: Right.
DESMOND: Right.
GROVE: And the PC upended the business architecture, to use that quote -- that phrase -- of the computer industry. And I think digitalization and digital distribution is going to change the order and, yes, there is going to be unhappiness just like there was unhappiness in the "computer industry" watching the new upstarts -- the minicomputer and personal computer makers -- take a lion's share of the revenue pie. But it happened and it's going to happen again.
LESSIG: Well, OK, let's remember an important moment in the explosion of the PC revolution. Everybody said IBM made such a terrible mistake in giving Microsoft the operating system and just licensing a version back. But IBM also had in its plan control of the ROM BIOS -- this was the startup chip that would make it so that it was a quote "IBM PC." It was Compaq that went and reverse-engineered that ROM BIOS to, then, establish the PC industry where there could be lots of competition among a lot of different producers all buying Intel chips but, still, lots of competition in the boxes that they produced that gave birth to the PC industry. Now, that reverse-engineering, under some views of intellectual property, is a crime, it was theft. It was theft to the IP that IBM had built into the original ROM BIOS. Now, it was because that view of theft wasn't permitted to capture the birth of the PC industry that we got the birth of the PC industry. And my concern is this 17
18 idea of theft will take over the lawmakers right now so that we won't get the equivalent of the reverse-engineering of the ROM BIOS that gave birth to the PC industry.
ARTHUR: Larry, if you could, if you had a magic wand and you could wave it and ...
DESMOND: I have one, actually ...
ARTHUR: Excellent. What incentives would you like to see or what legal, not too technical but what legal setup would you like to see in place where everybody could come out ahead? If it's true what I'm saying that there's a massive buildout ahead and there's a golden opportunity and a lot of productive growth available if only roadblocks can be taken out of the way, what incentives would you like to see for Hollywood, for Intel, for Silicon Valley, for content providers where everybody could be better off?
LESSIG: Well, my complaint is the set of roadblocks that are created by legally, government- backed monopolies.
ARTHUR: Right.
LESSIG: And so that's what IP is. Intellectual property is government-backed monopolies. And so I think the balance that we have to strike is one that doesn't allow these government-backed monopolies to be sources of concentrated monopoly power concentrated in these industries. So one reason we've had such an extraordinary concentration of intellectual property is intellectual property basically goes forever even though the framers of the Constitution said it was to be for quote "limited times." Eleven times in the last 40 years Congress has extended the term of existing copyrights basically guaranteeing to Hollywood that none of their work passes into the public domain. Now just think about this. Disney, the only ... you've all heard of the Grimm fairy tales and you sort of have this benign thought about the Grimm fairy tales. The Grimm fairy tales are awful, terrible, terrible stories. The sort of thing that you should never let your children see or read because ... brutal, moralistic ...
LESSIG: They're grim, right. They're grim tales.
LESSIG: But you think nice things about them because the Disney corporation took those stories and turned them into extraordinarily beautiful, for their time, retellings of the Grimm fairy-tale stories. They could do that because those fairy tales were in the public domain. Now, this company that made its, an important part of its past on taking stuff from the public domain and making great stuff out of it, believes in an IP regime that says, "Nothing we produce should ever be allowed to pass into the public domain." That there should be no Walt Disney that does to Disney what Disney did to the brothers Grimm. Now I just think that's a radical transformation of our past, of our tradition, and that we ought to have a much more restricted and, I think, very strong but limited set of intellectual property rights that give people incentives for a limited time and then their work passes into the public domain as it always has before the last 50 years.
ARTHUR: So 5 years instead of 95 years, or some such thing?
LESSIG: I actually propose a maximum of 75 years, but in 5-year terms to get renewed every 5 18
19 years. So if you don't want to renew it, it passes into the public domain. But this is quibbling about the details. The point is we've got to agree that public domain is an important part of building innovation and creativity, and that's what the past has demonstrated.
DESMOND: Would you not support, then, Jack Valenti's proposal to have copyright deals that can last forever and a day?
LESSIG: No, he said, "Forever minus a day."
DESMOND: Oh, sorry.
LESSIG: That was his -- when Mary Bono introduced the Sonny Bono Copyright Term Extension Act, she said we should perhaps consider -- because her lawyers told her perpetual terms would be illegal under the Constitution -- we should consider forever minus a day. No, I wouldn't support that proposal either. [Chuckle]
DESMOND: So how has this copyright regime been extended? How has it been distorted in the past couple of decades?
LESSIG: Well, you know, before the Internet there was very strong interest in intellectual property providers -- Hollywood and publishers -- who supported the idea of extending rights, and that made sense to them. And on the other side there was basically nobody. There is, in some weird sense, the public, but there was no strong interest on the other side. I think the Internet has changed that. It's changed it for two reasons. One is overly strong intellectual property protections -- as even Michael Powell, the head of FCC, suggested in a speech about broadband -- can actually chill technological innovation. So now, on the other side are very large industries that contribute actually much more to the United States economy than Hollywood does. And the other side of this is that there, because of digital tools many more people can be creators than at any time before. The cost of creativity drops dramatically. And so many more people are affected by this massive increase in regulation that has gone on in the name of intellectual property. So the Internet actually means more people are invested with finding a reason to strike a balance than existed before the Internet. So they got a free ride all the way up to now, but I think we're beginning to see a twisting back of that. In part, the only way we win is if it's a consumer revolution, but the only way we get a consumer revolution is if strong industries stand up to the Hollywood industries, and Intel was the first to do that in this hearing. But it will take more than Intel -- maybe not much more but more than Intel to fuel this revolution, I think.
DESMOND: Thank you, Larry. Let's take a few questions from the audience. Andy? With the lawyers is there a risk that the U.S. might lose its leadership in the information revolution?
GROVE: Oh, I think there's a very real risk, and it has nothing to do with the lawyers. It has more to do with the educational system in the United States as compared to its counterpart in different countries, particularly Asian countries. The rate of development in the technical fields in Asia, in Japan, in China, in Korea, in Taiwan, in Singapore is far faster than the rate of development in the United States. For a period of time we've actually lived off their education system with value added by our education -- our higher-education system on top of theirs. But, in the meanwhile, their education system has grown in capacity and capability both in terms of volume and in terms of 19
20 higher education, and native industries are developing very, very rapidly. And, probably, it's a fair approximation that two-thirds of personal computer development work -- personal computer design and development work -- is done in Taiwan and China today, and it's an export industry for them. You take that trend that has taken place in the last 10 years and project it for the next 10 years even without any mishaps on deployment, which could be related to what we've been talking on it so far, there's a real danger of continued shift of the center of gravity of technology, technological leadership, from the United States to the Asian countries.
DESMOND: I see. Second question. You alluded to this briefly. In terms of understanding the motivations of Hollywood, are they interested in protecting copyright because they're true believers or do they genuinely feel that there's a competitive threat which could undermine their power in the entertainment world?
GROVE: Well, I think both gentlemen on my side have alluded to the incumbents' natural tendency toward defending the existing business architecture -- I like that phrase -- architecture of business, business models, the rules and technologies that prevail to that. So whether we are talking of the original Luddites or we are talking of the people with the red flags in front of cars or we are talking of people who were fighting VHS, it is self-protection. I know how to do business with today's technology. I know how to do business in today's structure. I know who my competitors are, I play golf with them. I like my life, leave me alone.
GROVE: It's understandable and it has never worked for very long before. I mean, maybe it's worked for 10 years, maybe it's worked for 15 years. In the Internet time, it's going to work for 3 or 5 years but it's not going to work forever. Technology always wins in the end.
DESMOND: What will break the impasse? Considering what's happened in the courts lately, it seems that Hollywood's played the game pretty well.
GROVE: I don't know but my guess is that the continued erosion of the existing order is going to put pressure on the incumbent managements. How many years of sequential 5 percent revenue declines will the music industry take before they're going to scratch their heads and say, "You know, maybe we ought to get serious about digital distribution of music. That's what really people want." And they will discover what is obvious to the ponytail folks.
DESMOND: Brian?
ARTHUR: Well, if I recall right, the film industry opposed videos. This is a story that Larry knows extremely well. And when videos did come in by popular demand -- standard videos we go and rent -- not only did, were consumers better off but the film industry itself made enormous revenues out of videos. So I think that Hollywood, if we're talking about Hollywood as the antagonist here, doesn't realize that they stand to gain from the new technologies a great deal more than they stand to lose. And I think that the impasse could be broken if there was some way to demonstrate that to them. The video industry -- the film industry -- did extremely well out of videos that they had opposed. So I do think that we have to find some way to show Hollywood that they're not going to lose collectively.
GROVE: To illustrate the degree of opposition, the same Mr. Valenti that you've mentioned before 20
21 called videotapes the Boston Strangler of the movie industry.
LESSIG: Worse. It was a Japanese Boston Strangler because it was a Sony corporation that was ...
ARTHUR: ... Tokyo Strangler.
DESMOND: Tokyo Strangler. This is a question for Larry Lessig. What is the role of antitrust in this? Who stands to lose and gain the most?
LESSIG: Well, I thought that there was a useful role for antitrust to make sure that the platform for development and innovation stayed open and that no single company had the power to control the future of innovation on the PC platform or the Internet platform. And I thought we won that case, but apparently the current government thinks we lost that case, so maybe there is no future for antitrust in this particular context. But I thought it stood for a principle and the principle was in the context of technology justice, in the context of content, the dinosaurs shouldn't be able to call the shots in the future. And open platforms, which is what I think for most of the history of the Microsoft/Intel platform it was, an open platform; it supported lots of innovation without the platform owner controlling the future was the best way to get innovation. And I think to the extent that this lesson has been learned, the dangers in the future I don't think will come from companies like Microsoft who, I think, at least under one version of their next architecture, wants to develop a system which is another open platform that lots of Web services can be built on top of, not necessarily controlled by Microsoft. The real danger will be companies that are not building open platforms but closed platforms that try to marry content and conduit in a way which controls, gives them the power to control, how stuff is delivered or how innovation develops.
GROVE: Just so we give equal time, who do you happen to have in mind?
LESSIG: Well, I'm not yet convinced, but that's the form of businesses like AOL right now.
DESMOND: Very good. Well, there is a question here -- it's an open one ...
GROVE: For the record, AOL is ...
DESMOND: The owner of Business 2.0.
DESMOND: We believe in full disclosure at all times. Actually, it points to the next question. In the audience someone asked, What about AOL, what about Microsoft? How do these two companies play into the discussion we've been having today? I think you can probably all take a pass at this one. Andy, would you like to begin?
GROVE: The bulk of the discussion had to do with entertainment content, and while AOL now owns entertainment content ... But AOL is probably the most progressive on the subject for two reasons. And I'm not saying that because you are the host.
DESMOND: I would never imagine that.
GROVE: AOL, of course, is the epitome of an online company and an Internet company, so they 21
22 understand the power of digital distribution because they exist because of digital distribution. But even before that, Warner was the most progressive technologically, the most cooperative of the high-tech industries. To this day they continue to be. And without Warner, DVDs wouldn't exist today, I'm quite convinced. So there's a spectrum between the companies and there's more open to change and less open to change, and Warner is on the more open to change spectrum.
DESMOND: I'm glad you let me off the hook. Brian? This will be our last question, by the way.
ARTHUR: I don't have a comment on AOL specifically or Microsoft. I do think that what's extremely important is to keep pathways to innovation open. And some of the great companies like Microsoft have fought valiantly for that principle. The important thing in the United States is to keep innovation free and innovation open, and I think that should be the principle that policymakers have in mind in this area. DESMOND: Larry, are AOL and Microsoft on the right side of the equation here? LESSIG: They both could be on the right side of the equation -- AOL and Microsoft. We'll see whether both remain on the right side. DESMOND: What are you watching? LESSIG: I'll be watching AOL more closely right now, but I have great faith in your company here. DESMOND: Well, I'm sure they all read our magazine. CHAIR: Gentlemen. I hope you have all enjoyed tonight's program of the Commonwealth Club Silicon Valley and Business 2.0. Again, we would like to thank Brian Arthur, Citibank Professor at the Santa Fe Institute; Andy Grove, chairman of Intel; Lawrence Lessig, professor of law at Stanford University; Ned Desmond, editor and president at Business 2.0; the Tech Museum of Innovation; our audience here in San Jose; and those of you joining us on radio. And now this meeting of the Commonwealth Club Silicon Valley, part of the nation's largest and oldest public affairs forum, is adjourned.
22
23
Coming from Your Inner Self
Conversation with W. Brian Arthur Xerox Parc, Palo Alto, California, April 16, 1999 Joseph Jaworski, Gary Jusela, C. Otto Scharmer
Joe Jaworski: We wanted to meet and talk with you because of something you said: that for the big decisions in life you need to reach a deeper region of consciousness, that it takes courage to listen to your inner wisdom, but once you hear that wisdom, making decisions becomes very easy. We’re really interested in hearing your personal story as a way to begin.
I. Growing Up in Belfast
W. Brian Arthur: I grew up in Northern Ireland in Belfast. I went to university there, and grew up as a Catholic in a Protestant place. Being in the minority gives you a couple of things: one is the sense of being outside and a sense that there isn’t a place for you, and the other is that it makes you an observer. If there is not a direct path you tend to sit and watch for your opportunities.
I’ve got nothing against Belfast or Ireland. It’s got a dreadful reputation, but I find it a very warm- hearted place and have no regrets about coming from there. Life there is fairly tough and people are fairly tough and they have a very objective view of life.
My interests were in mathematics and engineering. I studied electrical engineering in Queens University in Belfast, and I entered University when I turned seventeen. It was ridiculously early. I wasn’t mature enough. It was a very narrow education. In the UK system, things narrow very early.
C. Otto Scharmer: What led you to choose engineering?
W. Brian Arthur: I was interested in mathematics and physics, but I wanted to do something more practical. I didn’t want to just be an egghead. I chose electrical engineering, but within a year or two I regretted it.
I was fifteen when I had to decide. My parents never went to university and nobody knew what engineering was. There was no career counseling. So I went there blind and discovered that I had no 23
24 interest in engineering and no talent for it. I was top of my class with first class honors but I had no real interest in it. Then I went to England for a year. It was a huge cultural change. After a year there I transferred to the University of Michigan.
JJ: How were you able to do that?
UC Berkeley, 1969-74
W. Brian Arthur: I went to one of my professors and said I wanted to study in the U.S. for a Ph.D. He wrote down all the contacts he had alphabetically, and Ann Arbor was at the top of the list, Berkeley was next, and so on. I was lazy. I applied only to Ann Arbor, got accepted, and went there to study mathematics.
I was there from 1967 to 1969. I didn’t like being in the Midwest so I transferred again, this time to the University of California at Berkeley. I arrived in Berkeley a week after People’s Park – talk about journeys. Belfast erupted in August 1969 when troops came in. The next month I went to Berkeley, which was also erupting. The five years I spent in Berkeley spanned the time from People’s Park to Nixon’s resignation. If I had chosen a five-year window from 1969 to 1974, I couldn’t have done better.
A few years after I had been in Ann Arbor, I was looking for a summer job in the States and I couldn’t get one because the Vietnam War was on and I was a foreigner. I had a professor who knew someone in McKinsey & Company. McKinsey hired me and sent me to Dusseldorf during the summers. I really loved that. The studies I was put on involved BASF, Volkswagen, Deutsche Bank, and a couple of others.
For me this was absolutely crucial. Imagine an education that consists of engineering, then mathematics, and then operations research, and then this counter-education in the background involving the political upheavals in Northern Ireland. I’d been brought up with a lot of political tension around, and I didn’t want to be part of it, but I found I couldn’t escape it in Berkeley. I wasn’t part of it, but I sure stood by and watched. There were many, many close incidents, tear gassings, and so on. This was very real for us at the time.
II. Learning the McKinsey Method: Strategic Cognition
Then the other thing that was happening was McKinsey. I learned more with McKinsey & Company than I did in graduate school. McKinsey was interested in large amorphous problems, like what should be done strategically. At that time, companies in Germany were still organized along divisional lines. McKinsey’s sophistication was very good.
In America, industry under Alfred P. Sloan’s ideas was reorganizing into profit centers, so this is really what we were selling. To McKinsey’s credit, it didn’t go in there and just reorganize on day one. They went into large companies like Deutsche Bank, or BASF, and they just sat and sat. They didn’t do anything. They just sat and observed and interviewed and observed and thought and went back and observed. It cost plenty to do this, but they were quite patient. This would go on for months until they had what I would now call a complex picture of what was going on. The opposite of that would be to come in with some cognitive picture saying, "You need to be reorganized this or that 24
25 way." They actually let a picture emerge, and this wasn’t lost on me. I would now call this an inductive rationality rather than deductive rationality. Rather than laying a framework on top, they simply let the framework emerge.
So strategic cognition was what McKinsey was good at. Somewhere along the line I absorbed all those lessons. They wanted me to come back and join in the firm in Germany and make my career there. In due course I probably could have been a partner, but I decided I wanted to go into science.
COS: What led you to do that?
W. Brian Arthur: I was studying operations research, and after McKinsey I lost faith in that too. Operations research was too mechanistic. I began to realize that the important things in business wouldn’t be decided mathematically. Operations research is good for scheduling fleets of trucks or production lines, but when it comes to something truly important a wider cognitive vision first makes sense before you make decisions. That couldn’t be done easily via the kinds of decision-making I was being taught.
To learn how to make sense of that, I transferred into economics at Berkeley. I was getting a Ph.D. in operations research, and I finished that, but I wanted to take a second one in economics. In those days Berkeley didn’t allow two Ph.D.s. I did everything but a dissertation in economics and they wouldn’t accept a second dissertation. They gave me a post-doc instead in the economics department. I’d been interested in the economics of third-world countries, particularly in the economics of population growth. I joined a small foundation in New York called The Population Council, on Park Avenue, and in due course I was sent to Bangladesh and then Kuwait and Syria.
Bangladesh: Doing the McKinsey
I discovered there that my scientific frameworks counted for nothing. I remember going down to the World Bank to the Bangladesh desk and coming away absolutely, totally unimpressed. [I asked,] "What’s literacy like in Bangladesh?" Well, 23 percent are illiterate in this region and it rises as high as 27 percent. I said, "What does literacy mean?" They didn’t know. I said, "In most Moslem countries it means that you can read a passage out of the Koran, and of course people may have memorized it. What about real education?" Well, they didn’t know.
So I did a McKinsey in Bangladesh. I just sat there for a long, long time and did nothing until the whole goals and structure really surfaced. I was with a superb sociologist colleague, Geoffrey McNicoll, and we both did interviews, and thought, and sat on the edge of desks, and I reasoned that if this was happening, that must be happening. If that’s happening, it might imply such and such. We were interested in landlessness and the incentives for child-bearing and so on. We cognized or structured that whole thing. We wrote an 80-page paper that later came to be seen as a classic on Bangladesh.
I got bored with that after a couple of years, and went to work in Vienna for five or six years.
Vienna, 1977-82
COS: When was that? 25
26
W. Brian Arthur: Nineteen-seventy-seven to 1982. In 1981 I was thirty-five, and I wanted to find a job in academia. I knew there was a chair becoming vacant in Stanford in population studies and economics, so I came to Stanford as a one-year visitor. I was hoping to get that chair, and indeed I did in 1982. I was in the department that dealt with third-world countries, the Food Research Institute. I had another appointment in economics. I had a chair in population studies and economics, the Morrison Chair. I was thirty-six and I had a professorship in human biology, not that I knew any biology, but I taught kids out of that department. That went on for a long time. I resigned that position in 1996. There’s quite a long story behind it, but really it amounted to, as you were saying before, reinventing yourself.
JJ: Was there any sort of triggering? Was there any triggering mechanism at that point that caused you to want to reinvent yourself?
III. The Path to the Big Idea
W. Brian Arthur: In 1979 I started to read a lot of molecular biology. I was very influenced by a book called The Eighth Day of Creation by Horace Freeland Judson. It’s a history of the discovery of the structure of DNA, and the discovery of how the genetic code worked, and the discovery of the structure of the hemoglobin molecule. I was fascinated. I started to read about enzyme reactions and the writings of Jacques Monod, a French molecular biologist and Nobel Prize winner. He’d written a book called Chance and Necessity where small events could get magnified by positive feedbacks and lead to different enzyme reaction paths. I began to realize that the counterpart in economics to positive feedback was increasing returns. I started reading the physics of positive feedback, and particularly the work of the German Hermann Haken at Stuttgart and the Belgian Ilya Prigogine, a man I am very fond of.
I realized that positive feedback in economics had to do with increasing returns. I realized that economists couldn’t deal with increasing returns because they led to multiple possible outcomes. Basically, if there are several competing things and one tends to get ahead, it gets further ahead. Two hundred years ago the languages of Central Europe were French, English, and German. The more one was adopted the more useful it was to adopt that language, and given different set of historical events – say the victory of Napoleon overall – we might have all had to speak French internationally. There are several possible outcomes and historical accidents led us into the gravitational orbit of one of them. I began to see mechanisms of positive feedback causing competing bandwagons where in time, one of several competitors could really start to take over and lock in.
I realized that the economists had avoided this phenomenon of increasing returns because they didn’t like multiple outcomes. Schumpeter had said that unless an economics problem leads to a unique equilibrium outcome, it’s chaos that is not fully under analytical control. And he was very upset about it. I realized that what we had to do was to allow for the possibility of multiple outcomes. What I contributed was to map those sorts of problems into stochastic processes that were nonlinear. I started to do a lot of work on nonlinear stochastic processes with some Russian probability theorists. This was June 1979.
COS: Can you describe the context of that particular moment when you got this intuition?
IV. The Moment of Epiphany 26
27
W. Brian Arthur: I’d taken a two-month leave to be in Hawaii. My wife had just finished her Ph.D. and I went to Hawaii. I was reading a great deal of molecular biology. Normally I read physics. If there was a moment of epiphany, it was in June 1979 when I read a little essay that Prigogine had written. I forget what he called it, but it covered everything from the way termites build nests to the phenomenon of languages taking over, but it was about positive feedback, and instantaneously I realized I had something that was important in economics. All I needed to do was figure out how positive feedbacks worked in economics, and it took another ten years to do that. But suddenly, within about two or three weeks, everything in economics fell into place for me. It was a period of very, very intense intellectual excitement.
JJ: Did this germinate in Hawaii when you were out there?
W. Brian Arthur: No, but the scene had been set in Hawaii. That’s where I had been reading all the molecular stuff. It was about two months after that; I’d been totally primed. I think I read Monod’s book, Chance and Necessity. A lot of things happen by chance. Some things happen by necessity – deterministically – but with positive feedback, the necessity magnifies the chance. It locks things in. I began to get a very good feel for positive feedbacks, and I realized that the counterpart in economics was that small events can lead the economy to quite different structures. You know, if Napoleon hadn’t done such and such on the morning of Waterloo maybe, dot, dot, dot.
I realized what was missing was the methodology, and I knew where to get it. I had to delve deep into nonlinear probability theory. It took me years to do this and I’m still not very good at it. I’m not a professional probability theorist, but I had to get good enough to work on a level with real professionals.
Big Idea = Big Threat
The other thing I realized was that when I floated this around, economists got threatened and I got scared. I had the ball and the net was wide open. I was goal shy because I knew there would be hell to pay. I was saying that small events can lock the economy into different structures and that it’s fractal – that there are structures within structures, that the entire economy isn’t the best of all possible worlds. Capitalism does not lead you to the best of all possible worlds. This was regarded as a supreme threat. It was the middle of the cold war. There was a lot of ideology – it was the Reagan- Thatcher years – and the whole edifice that had been built up for 200 years was threatened. You couldn’t do economics statically anymore. The equilibria that manifested were not the best of all possible worlds. Markets were not perfect. Small events could lead you to inferior solutions, and I knew there would be hell to pay. What I didn’t realize was just how much hell had to be paid. So I wrote this up, waited for two or three years, and then wrote it up [again] in 1982. I couldn’t get the articles published.
JJ: After you wrote this, you couldn’t get any of the articles published?
W. Brian Arthur: No. In the first ten years of my career in economics, I published many articles and got a chair at Stanford. During the second ten years I published one article. In the end, that led me to leaving Stanford.
27
28
JJ: But this is the story of any entrepreneur, you know, somebody who is willing to go ahead and take a stand and then they pay hell.
How Big Ideas Happen: An Archetypal Journey
W. Brian Arthur: This is true, but I didn’t know this at the time. I’ve read a lot more history of science. I wish people would tell you in high school. You think somebody gets an idea in their bathtub or the shower and everybody realizes it’s valid immediately and you’re led on people’s shoulders down the street with the crowd cheering. It’s not at all true. So I couldn’t publish and I spent the next ten years in a professional hell, but I kept writing. The support I got was always at the very top.
JJ: And where was the article finally published?
W. Brian Arthur: The Economic Journal in England, but I sent it to the American Economic Review and the Quarterly Journal of Economics. They couldn’t find any technical fault, you know, why this isn’t economics or whatever. It was a horrible period.
COS: What kept you going? When everything was turning against you, why did you continue on?
W. Brian Arthur: I thought it was deep and I thought it was right and I thought it was about the deepest thing that had happened in economics in a good while, and I still think so. It’s a totally different way to look at things. That kept me going. I also had support from Kenneth Arrow at Stanford and a few [other] people. Arrow gave me a Guggenheim Fellowship in 1987.
I had support from Tjalling Koopmans and Kenneth Arrow and a layer of Nobel Prize winners. They were open-minded. It was one layer down that was the problem. I’ve no regrets, but it was a very horrible period of getting rejections back and being regarded as a charlatan and a pariah. At the end of my years at Stanford I was probably less employable than I was at the start. You know, to this day, I’m not sure a respectable economics department would have me because there is always a minority that can block an appointment. This will not be true forever, but it is true now.
V. Santa Fe Institute
Arrow brought me to the Santa Fe Institute in 1987. And Philip Anderson, who is a physics Nobel Prize winner, sprinkled holy water on all the ideas.
JJ: That was in the early days of the Institute.
W. Brian Arthur: Yes, very early. They hadn’t really started. It was August 1987, and I gave a talk there in front of twenty people. It was a who’s who of economics and physics. The economists bristled at what I was saying. The physicists just sat there nodding. Phil Anderson said, "I thought economics was dull and boring, but this stuff speaks to us." So after that I was brought back to the Santa Fe Institute in 1988 to start their first research program. I was backed by John Reed from Citibank, who gave us a lot of money. There was no other research at SFI, so I set up the first research program and was able to set a lot of the style of how things were done there. No departments, no students, and a very open atmosphere of just throwing out ideas and going to the fundamentals.
28
29
John Reed was saying, "Do anything you want, just don’t be conservative." Arrow and Anderson also said do anything you want. I said "What precisely do you mean?" "Go deep into the foundations and change anything you want."
JJ: So this was a brand new opening at the Santa Fe Institute?
W. Brian Arthur: Yes, and that saved me. I set up the first research program there. George Cowan, president of SFI, took a huge gamble on me. Bob Solow and others at MIT told Cowan he had made a dreadful mistake, that I was nobody. George backed me anyway. I could pick anyone I wanted to bring to Santa Fe, so I brought John Holland and Arrow and Frank Hahn and others. These people were heavyweights and I was able to put the problem to them: If you had to rethink economics, what would it look like? And that is what we did. I don’t know how it will shake out. This is what I’m good at – to take something amorphous and figure out how it works.
The Inner Journey
I think these are patterns that are common to many people’s lives, especially if you’re a bit older. In the atmosphere of the ’80s I had everything anyone could want, but I was profoundly unhappy. I wanted to do this sort of work and nobody would publish it and listen.
Then in October of 1986, my wife, who has a Ph.D. from Princeton in mathematical statistics, got deeply involved in Buddhism. I couldn’t make head or tails of it, but I was interested. I’d been brought up Catholic. I had renounced that in my early twenties. I became agnostic and had a kind of scientific- engineering agnosticism. In 1975 that started to change. We traveled through India and Katmandu. Suddenly I began to realize that there was a dimension out there that was trying to shout to me but I’d nothing to hear it with. I started to read Carlos Castaneda’s books. Like a lot of people in those days, not only did I read them, but I reread them and reread them. I don’t know where they are now, but they’re all marked up and absorbed and digested.
VI. Taoism and Economics: Mechanical Order vs. Unfolding
COS: How do Buddhism and Taoism relate to economics?
W. Brian Arthur: Standard economics is very good for being shoehorned into an image of 19th- century physics. It was precise and accurate and static; it concerns itself with equilibrium. I began to realize that what really interested me was to see the economy not as static but as unfolding, and as patterns that were always unfolding.
I began to realize that if patterns were always unfolding it gives you two questions or problems. The economy is always unfolding, and at a more fine level business is always unfolding. John Seely Brown says if you leave your job for a couple of weeks and come back, the whole atmosphere is different. He’s exaggerating, but you know the game has changed. So let me try and contrast that with a different view. The standard way of looking at cognition and decision-making is very different from this other view I stumbled upon.
You were asking how that fits. If you ask Taoists how they see the world, the first thing they’ll tell you is that the world is changing. Everything is always changing, everything is always unfolding, 29
30 and it is our job as human beings to allow things to unfold. You can give a little nudge here and a nudge there, influencing things at the proper time in your own way, but the world is not seen as a machine. The world is seen organically as a collection of unfolding patterns. When I worked on my economic increasing returns theories, before I studied Taoism, I gave a talk at the University of Hawaii in 1985 and a student from the Chinese mainland came up to me and said, "All that you say has been said before." And I said, "All right, give me a citation." He said, "It was all said by Lao Tzu." I said, "In that case, I’m honored."
Taoists see the world as patterns that are unfolding. I’ve gone back and read Sung-Dynasty Taoism and Neo-Confucianism. Cheng I, and Cheng Ming Tao, and various others writing and teaching in the late 1000s. It’s remarkably contemporary. They taught that all was in flux but that everything structured itself according to inner principles that governed it. Now we’d call those laws. They said principle is one, but its manifestations are many. In other words, things in this world emerge from elements that structure themselves. The mind, they said, is not a vessel to be filled with facts or ideas. It too emerges. The mind is an emergent phenomenon. All this they said a thousand years ago.
Complexity Theory
Let me talk for a moment about complexity theory. It’s really a movement of the sciences. Standard sciences tend to see the world as mechanistic. That sort of science puts things under a finer and finer microscope. In biology the investigations go from classifying organisms to functions of organisms, then organs themselves, then cells, and then organelles, right down to protein and enzymes, metabolic pathways, and DNA. This is finer and finer reductionist thinking.
The movement that started complexity looks in the other direction. It’s asking, how do things assemble themselves? How do patterns emerge from these interacting elements? Complexity is looking at interacting elements and asking how they form patterns and how the patterns unfold. It’s important to point out that the patterns may never be finished. They’re open-ended. In standard science this hit some things that most scientists have a negative reaction to. Science doesn’t like perpetual novelty. I once asked John Holland, who’s knowledgeable about chess playing, if chess has reached some sort of equilibrium where if everyone plays their best, games are lost and won, but chess overall does not progress. He said no. There is novelty in what’s discovered century by century in chess. A good tournament master now could possibly beat a grand master of a hundred years ago because the envelope has been pushed out of what’s known. So anything complicated and interactive seems to unfold and develop new structures.
The Mechanistic View of the Old Economy
Now switch to business or the economy. The old thinking is that business and the economy are mechanistic. People talk of linkages, that things have to be "on the right track," that we need to fine- tune things, get it up to speed. If only we understood the mechanisms, we could fine-tune the economy. At deeper levels in business there are decision-makers, agents, and at any time each agent faces a set of problems, probably with a capital "P," and to those problems there are Solutions. This just happens to be a structure we laid on business, trying to make it a science. We believe there are Problems and there are Solutions. Implicitly it means that if you are managing there is a feeling here that you can actually frame the problem correctly so that there is a Solution with a capital "S," and it’s up to you to learn how to arrive at that solution. But all this only works in repetitive business, where you can 30
31 optimize and the problems are well defined. It appears in that case that management’s problem is to optimize, to get it right. Lower costs, get quality up, keep everything moving, make it smooth, make things reliable, solve the problems, and find solutions. That’s old thinking.
The World of the High Tech Economy
Let me contrast that with high tech. This is an article I wrote in the Harvard Business Review a couple of years ago, August 1996. There are several things that are different about high tech. One of them is that there are typically increasing returns, network effects, and upfront costs. So if you are the people who invented Java, you can make billions if you can lock it in. If you can lock people into doing documents digitally by Xerox rather than Canon, you can take most of the market. This is typical. High tech in my metaphor tends to be much more like a casino. It’s not the halls of production in repetitive industry, but rather the casino of technology. There are many tables in this casino. And at each table are different games. At this table we’re going to start up digital banking, and the outcome of the game is that two or three of the key players are going to take 90-something percent of the market. So this is not a situation where everyone gets 10 or 15 percent market share. You typically find 80 percent market shares, 70 or 80 percent, like CompuServe or Microsoft have in their markets. The next player might have 20 or 30 percent, and then there are a few bit players. This is because there are increasing returns and diminishing costs, and the more advantage you have the more advantage you get. The more people who use Windows, the more likely I am to use Windows. But it could have been some other operating system. When there is competition in this area, through those sorts of network effects or upfront cost effects, a winner will lock in most of the market. With Windows 98, the first disk will cost me maybe half a billion dollars and the second will cost me two or three cents. So the more I put out the cheaper my unit costs are. Therefore, the more market I take the more cost advantage I have. The more people are using that then the more that creates a network of users and so on.
Winners Take Most
So in competitions in those areas, the winner takes most. This sets up a totally different set of problems for management. The problems are not optimizing ones. You’re sitting at this table and starting up in digital banking. You don’t know what the technology is going to be. You don’t know who’s going to sit at the table so you don’t know who the competition is going to be. You don’t know what the government regulations are going to be. You don’t know how the technology will work out. You don’t know how it’s going to be received by the consumer, and you don’t know whether it will work. You don’t know how consumers will take to it. You don’t know what ancillary technologies are going to be used and what alliances people are going to cobble together to make it all work.
And so, the point I want to make is that there is no well defined problem with a capital "P." Imagine it’s five years ago and you’re thinking about Bosnia. You’re in the State Department or the UN and you’re going to put in a UN peacekeeping force. If I said to you, "Optimize the problem in Bosnia," you’d say, "What problem?" I mean, clearly things aren’t right, but is there a correct problem? Does a problem exist? No. I don’t mean there is no problem, but there is no correct problem. All you can say is that you have this situation and there are many ways to cognize it.
These situations have a Rorschach inkblot feel to them. You have to read the problem into the situation. If I’m sitting here in Silicon Valley, it’s not that at any day I might have a particular problem 31
32 with a capital "P." My supplier may not have sent me whatever processors I need or backup memory. That’s a problem, and maybe you can do something about it. But what I’m facing more typically as an entrepreneur or business person is a set of situations. And what I’m trying to do is to make sense of these. John Seely Brown says the challenge in the old economy is to make product. The challenge in the digital economy is to make sense.
So in a sense there isn’t a problem; there is a situation or a set of situations and they continue to unfold. And your job, should you accept it, is to make sense out of it.
It’s like your own life. If I said to you, what’s your problem? You’d say, I don’t know. Is there a correct problem? No there isn’t. You could get banks of therapists and go deeper and deeper. So you have a set of situations you’re facing. And there might be less appropriate or more appropriate means of dealing with these.
VII. The First Thing You Do Is Observe
W. Brian Arthur: You can approach this from the point of view of complexity or economics or Taoism; certainly try all three.
From the economics side, think of your own life or Bosnia or Belfast or high tech: You’re in a situation and varying conditions can be put in that.
Cognition is never extracted from the situation. You don’t make sense from the situation, you impose sense upon the situation. Confusion is the absence of the framework, and known confusion just means that you have framework. You can label it. We have a nice saying in Belfast, "If you are not confused, you don’t understand anything."
So what is facing management in high tech is confusion. The job of management in high tech, at the highest levels, is not to manage but to find frameworks. Once you have frameworks you’re willing to impose, they imply the appropriate reactions. Not optimal reactions, but appropriate. So you can’t optimize in this area. All you can do is to act appropriately.
Switch then to Taoism. Taoism keeps saying that the world unfolds. There is no truth. There is only that which you impose upon it, and you can’t move the world. But you can move yourself appropriately. So the way Taoism would inform martial arts is to say you don’t know what your opponents are going to do, but when your opponent moves you can react appropriately.
So you don’t have to face 4,000 pounds coming at you head-on. You should be able to move to the side and deflect it. This way of thinking would say that there is no correct solution. You allow the world to unfold and you act appropriately.
So I’m sure you’re beginning to see why I say the first thing you do is observe.
JJ: It’s totally, completely understood.
W. Brian Arthur: Okay. Or think of driving at night. You’re driving on a narrow road in New England late at night with your headlights on. There might be deer or animals on the road. It has just 32
33 snowed and there may be a little sheet ice out there, you’re not sure. What problem are you solving? There isn’t a problem. You’re not optimizing anything. You’re actually cognizing your feeling, you’re working from here, not from the head. Moment to moment you’re recognizing what you’re in and saying, "It felt like black ice going around that last corner and that wasn’t true half an hour ago, therefore I should do such and such." You are conforming to what’s arising, but not in a passive way. You are pre-positioning yourself. So in that sense it’s all about the position.
This is the case with Taoism. Taoism doesn’t tell you to optimize your life. It doesn’t tell you that life is going to be happy. It just tells you that the world is always changing. It doesn’t tell you to mutely conform in a passive way, but you shouldn’t struggle either. You just go with whatever it is, and at each stage you put your all in appropriately.
Complexity
The third thing is complexity. When you start to take the world seriously from the complexity point of view, you begin to see that organisms are in a world of continual change. If you look at a species in an ecosystem, the ecosystem is much more changeable than we ever supposed. I was watching "Nova" a week ago and they were saying that we’re in a period of remarkable stability. This wasn’t true ten thousand years ago, when there were huge changes worldwide. So it’s hard to talk about optimization or that a species is optimized. Everything is mutually adapting to everything else all the time, and the complexity point of view asks how things adapt to various circumstances.
As we shift into a high-tech economy, making sense of situations and then taking appropriate proactive kinds of actions counts. This is opposed to the pre-high-tech economy, where the situation was unchanging, problems could be well defined, and therefore optimal solutions were possible.
A totally different set of rules apply in this new environment. You hang back, you observe. You’re more like a surfer, or a really good driver at night. You don’t act out of deduction; if you do, you’re going to get creamed. You’re acting out of an inner feel, making sense as you go. You’re not even thinking. You’re at one with the situation.
Taoism
Let me go back again to oriental thinking, to Taoism. In oriental thinking, you might just sit and observe and observe – and then suddenly do what’s appropriate. You act from your inner self. Traditionally, Chinese and Japanese artists sit and look at a landscape. They’ll sit on a ledge with lanterns for a whole week just looking, and then suddenly say "oooohh" and paint something very quickly. Consider a tea ceremony. With deep training and deep observation, you’re reacting appropriately, and the appropriateness of the reaction depends upon the degree to which you are at one with the situation. It’s the same with martial arts: If you have to think in martial arts, you’re dead. The twenty or thirty years of training you’ve had mean that you’ve internalized lots of possible patterns and instinctively know how to react.
VIII. Management in the High Tech Economy
This has a lot of implications for management because it’s saying that what counts is where you’re coming from in your inner self. Now imagine you take a CEO out of the processing industries who has 33
34 optimized Pepsi-Cola, for example, and you put him in Apple Computer. What’s he going to do? He will bring that one sort of cognition, "cost down, quality up" or whatever the mantra is, with him and that’s it.
Now think about Steve Jobs and Bill Gates. When Jobs came back to Apple he said, "Well, the reality is the Internet now. What are we going to do about it?" He turned Apple around. Gates and Jobs are good at very different things, but they both know how to distance themselves from the "problem" and expose themselves to something different. You wait and wait and let this experience well up into something appropriate. In a sense, there is no decision-making. What to do just becomes obvious. You can’t rush it. Much of it depends on where you’re coming from and who you are as a person.
JJ: Right. I totally understand.
W. Brian Arthur: It’s true of first-rate scientists also. The merely good scientists are able to take existing frameworks and overlay them on to some situation. The first-rate ones just sit back and allow the appropriate structure to form. My observation is that they have no more intelligence than the good scientists do, but they do have this other ability, and that makes all the difference.
You can see this in business as well. Sam Walton didn’t just use some old framework to create Wal- Mart. He said "Well, hang on here. We’ve got computers, we’ve got inventory, and we’ve got a whole network, so what are we really trying to do?" It may take months or a year to figure that out. You cognize things or piece together the framework from a different perspective, and if it is appropriate then it works.
This has very practical implications. Suppose I’m a large New York bank and I want to set up digital banking. The worst thing I could do is to bring in somebody who has been successful from the processing industries. They’d come in day one with a ready-made framework and say, "Okay, we need such and such and this is how it is going to work and it’s going to be like this and such."
That’s appropriate if the problem is conventional, but it would be better to get someone who says, "Hmmm, digital banking, we’ve never seen anything like it. What’s really going to count here. Who is going to be in it? What will consumers accept? What is the government going to do? Is this when you take all or nothing?" If that person is any good, they’ll be acquisitive and take a piece from here and a piece from there. They’ll think, "The early days of such and such worked this way. If we look at digital banking like software, we can expect this. If we look at it as X, this might happen." Cognitions are built up piece by piece from the first principles.
JJ: Can you help me understand what you mean by "cognitions"?
Cognition
W. Brian Arthur: I view cognition as a framework that one would associate with something – for example, if you see someone and all you are seeing is pixels upside down in your brain. But the brain is structured so that certain features of the pixels trigger an association, and the association starts to trigger certain labels and you say, "Oh, that’s old Fred, my friend," or somebody. That’s a cognition. You’ve brought a set of associations to an amorphous pattern, and those associations can be quite metaphorical: "Oh this is just like Munich in 1936." It can be almost primitive associations: "This is my 34
35 dog. It’s not just that fuzzy set of brown pixels that are in my brain. This is a dog, and moreover I know it’s my dog and I saw it this morning." We do that with various degrees of sureness.
I don’t know if you remember those 3D patterns, where if you stare at them all you see is patterns, and then suddenly if you stare long enough, a three dimensional aspect comes out. Now what’s happened cognitively? The three-dimensional aspect didn’t come out of the paper. Certain physical things happened within certain sub-features inside your neurosystem that imposed an understanding of three dimensions, or a framework of three dimensions, on that paper. Moreover, it fits because if the 3-D was random it wouldn’t jump out at you. There is a consistency to the pattern. It looks like dancing horses or something. It fits with other associations. The horses look like horses. So cognition is an imposing of understandings and an imposing of previous associations. You know it is cognition when you hear someone use the phrase "it’s like." So if I describe something, I’m asking for a cognition, "Well, it’s like this; or, here’s what he’s like." You are searching for all the associations.
IX. Accessing the Deeper Levels of Knowledge and Knowing
COS: When you made the distinction between the good scientist and the first-rate scientist, you said the good scientist takes an issue or a problem or a situation and then imposes a framework on it. Then you said the first-rate scientist would do something else. They would expose themselves to the situation and wait until the pattern emerges. So what’s the difference? Isn’t that a different type of cognition than what you just described? Where does that come from?
W. Brian Arthur: This is a good question. A good scientist says, for instance, in economics, "This is the principal-agent problem, this is a game-theory problem, this is an overlapping-generations problem." In fact, graduate studies in any of the sciences and economics consist of teaching students thirty to fifty different cognitions until they’ve mastered them. Then they learn to apply that, and later, when they’re working for the World Bank, they can say, "Okay, it’s like a principal-agent problem." If you get good enough at this, you actually forget the labels and internalize it. You just have an inner feeling. I’m not saying that’s bad. That’s done, and there is an appropriateness to it.
First-rate people are saying something deeper. They wait and they don’t say anything. They go as deep as they can. They’re examining fundamental beliefs and saying, "Well, I could really move this company as a shrink-wrapped application, but it needs a lot of change." A first-rate person says, "Is that what we need to do? Is that the business anymore? Hasn’t it changed, and aren’t we really going to get into such and such?"
The way you can tell what the difference is that in the first type of cognition, the new business world becomes an appendage of the old. Encyclopedia Britannica CD becomes an appendage of Encyclopedia Britannica leather-bound volumes. But if you wanted to cognize that business appropriately in 1991, you’d have said "Encyclopedia content is just software. It will all be available on CD. Volumes are nice for libraries and people who like to have that around their house, but the real business is going to be selling these things and that should be done separately. The leather-bound books ought to be an appendage." It’s like a gestalt switch.
First-rate people go to the root. They don’t just ask what’s appropriate or what appropriate framework can I dig up and impose. They will study the situation from many, many angles and then say
35
36 fundamentally what really is going on here. They may borrow ideas and cobble together a very different framework.
There are many types of understanding. It may be that there is an overall simple type of understanding where you just say, "Oh, they’ve got an inventory problem here." Then there is the deeper kind of understanding I just talked about that asks, "What really is the problem here?" The first type of understanding tends to be the standard cognitive kind that you can work with in your conscious mind. But there is a deeper level. Instead of an understanding, I would call this deeper level a "knowing."
JJ: Yes.
Sources of Inner Knowing
W. Brian Arthur: The inner knowing comes from here [pointing to his heart]. Every one of us has experienced this in different ways. You may not be conscious of it, but you have.
I remember being on the side of a mountain one time in Santa Fe. The sun was setting and I was a third of the way up and I thought, "It’s time to get off and come down. I don’t want to break an ankle or something." But then I had a feeling that I ought to go on up the mountain. The sun set, and the moon rose, and it was dark, and I couldn’t see the path. I had been doing a lot of training at that time, and I think I had a knowing that I should go up the mountain. Within that, I had a knowing that if I couldn’t see the path it was likely I would kill myself. There was quite a sharp fall-off at the side of the path. But each time I didn’t know where to go, all I had to do was stop and ask here [pointing to his heart]. That was a knowing, and it was never wrong.
JJ: I totally understand.
W. Brian Arthur: I don’t know or can’t say where these knowings come from, but I tend to go back to oriental thinking. The Japanese and Chinese would say that everybody has these abilities, and the reason that a Japanese management trainee would go off and do Zen training is because that is considered to be cultivating yourself as a human being. One practical reason is that it allows you to get in touch with your inner self, this deep inner knowing. This is something the West doesn’t have much patience for. All I can say is this: When it is a knowing, you really know and it’s a total conviction. It doesn’t mean that I know everything that is going to happen or I know what mutual fund to put my money in, though sometimes I do get convictions that seem to be deep, and you know, you always ignore them.
Suppose I was parachuted into some situation in Silicon Valley. It’s not a problem, it’s just a situation that is complicated and changing and unfolding. I’m trying to figure things out. I would observe and observe and observe and then just retreat. If I’m lucky, I would get in touch with some deep inner place and then allow that knowing to emerge.
JJ: You know, that’s what David Bohm told me. He told me that this is the way to operate.
Coyote Café, Santa Fe
36
37
W. Brian Arthur: Yes. In a sense there is no other way to operate. One time Phil Anderson and I were sitting around in the Coyote Café in Santa Fe and somebody says, "Hey Phil, you play chess?" "No." "Do you play checkers?" "No." "So you’re not any good at that sort of thing?" and Phil says, "I play Go." "Oh," I said, "Phil are you any good at Go?" And he says, "Yeah, I’m not bad at it." "How good are you Phil?" "Well," he says, "there are four people in Japan that can beat me." So you say, "Why did they beat you Phil?" He says, "They meditate." And I told him "So, they’re coming from an inner place," that inner place of knowing. But you see I’ve run out of vocabulary.
JJ: There’s no way to give voice to it. I just wanted to hear you try to give voice to it because I believe your whole experience was saying that to me, and that’s why I asked the question about the cognition. There is another place . . .
W. Brian Arthur: Well, there is another place, and I think that place is to retreat deep, deep, deep into essence and then let things emerge. The distinction people often make is that this is not doing, it’s being, and our culture here is one of doing. The question is not "What are the appropriate actions?" but "What is the appropriate being?"
JJ: Exactly. This is the true point. This is why we came to see you. That’s exactly right. It’s all about being. Leadership in business is not about doing, it’s more about being.
W. Brian Arthur: It’s an expression of inner being. I’m just amazed sometimes. I bought my daughter an iMac and I have a G3 myself – they’re both Apple computers. I look at the select feature and I think, "This works beautifully. Why does it work?" Like a lot of people, I have had many computers. But what is it that’s so nice about the G3? There’s something lovely about it as a piece of technology. It actually doesn’t have to do with decisions; it has to do with the inner being of Steven Jobs. There’s an elegance to anything he touches, and that’s a function of being. It’s what he’s about. What he’s about is technology. Gates is not about technology, so anything that comes out of Microsoft is usually mediocre in terms of technology. Gates is about something different.
Listening to the Inner Place Where Knowing Comes to the Surface
It’s hard to find language for this. Someone like Csikszentmihaly talks about getting the flow, and flow has to do with things just unfolding, with listening to that inner place where knowing comes to the surface. It’s like when I knew I had to leave Stanford. I had an inner knowing. Something tells you that you ought to do such and such.
Gary Jusela: There is something that occurs to me about some of these stories you’ve been telling that’s a paradox to me. You talked about what might happen in leadership development in Japan. They go off and do this deep Zen-like meditation or actual real Zen meditation. What you’re describing in that sense of going into your essence runs so counter to our culture here in the U.S., and yet there is this pattern-recognition thing going on here and not going in Japan it seems, economically in terms of finding this new chapter. Doesn’t that seem paradoxical? What’s going on? They have the means. Japan is coming out of this deep oriental tradition along the lines of what you’re describing. This place is like pure rational science in a lot of ways, and yet there also is some other overlay. So how do you explain that?
Two Levels of Knowing 37
38
W. Brian Arthur: If you get to a very deep level of knowing, like some high Rinpoche or Taoist master, it doesn’t mean your knowing is about what would be immediately applicable in a modern world. If it did, the Dalai Lama could set up a lot of high-tech companies and make a lot of money. Again, I’m in waters here that I don’t understand or normally talk about.
Level 1 of Knowing: Knowingness Arises from Immersion
Maybe there are two different levels, but one deep level is when you have been so immersed in an area that it informs every fiber of your being – martial artists or oriental artists strive for this. They absorb and absorb and absorb and then they forget, and it becomes an inner part of their being and the knowingness can arise out of there. That’s one level.
Level 2 of Deep Knowing: Knowingness Arises from Grace
I suspect there is an even deeper level that the West might call grace. I don’t know what you’d call it, but these knowings don’t necessarily have to come out of years and years of previous exposure.
So to come back to your question, I think the Japanese have a deep, deep inner cultural exposure to precision. They had it with ceramics. They had it with the types of art that were done, and so it was natural when the transistor came along and high-precision optics and high-precision plastics and high- precision processing in manufacturing, that they would be very good at that because they had centuries of that kind of cultural knowing. It doesn’t mean that they’re good at what counts here, which is combining different elements into something else. That’s a very different set of skills. And again, I’m not sure how that comes out of the California culture, but it’s certainly here.
GJ: Right. There is something spiritual going on here, in terms of a level of spiritual knowing, that’s not just pure "traditional" science. You’re talking about a different kind of science here.
W. Brian Arthur: Yes. If I did get the right answer and I’m sitting in a Zendo in Japan and it’s "Ahhh, yes, Internet company of Zen, aaahh." And you leap outside and throw away your Zen stuff and where are you going to get the infrastructure, where are the venture capitalists? You’re right by Sand Hill Road here. Where are the patent lawyers [in Japan]? You can do it [there], but immediately the third parties would cognize it into what they deal with in Japan. They’d say, "Oh, you’re trying to do such and such. That’s just like microprocessing of such and such a kind and we’ve done that." Immediately they’ll shoehorn you. So what’s going on here is a subconsciously understood culture set of reactions so that if you do have an idea, you will have an appropriate infrastructure for that idea.
X. Sensing What Is Wanting To Emerge In the World
JJ: I want to tell you that the things you’re saying are really profound. I was just so struck by what you said because Martin Buber, the existentialist philosopher, writes about it. He said:
Free is the man that wills without caprice. He believes in the actual, which is to say: he believes in the real association of the real duality, I and You. He believes in destiny and also that it needs him. It does not lead him, it waits for him. He must proceed toward it without knowing where it waits for him. He must go forth with his whole being: that he knows. It will not turn out the way his resolve intended it; but what he wants to come 38
39
will come only if he resolves to do that which he can will. He must sacrifice his little will, which is unfree and ruled by things and drives, for his great will that moves away from being determined to find destiny. Now he no longer interferes, nor does he merely allow things to happen. He listens to what grows, to the way of Being in the world, not in order to be carried along by it but rather in order to actualize it in the manner in which it, needing him, wants to be actualized by him – with human spirit and human deed, with human life and human death. He believes, I said; but this implies: he encounters. (Martin Buber, I and Thou [trans. Walter Kaufman] (New York: Macmillan, 1974)
What I was hearing when we were talking is that the work is to sense what it is that is wanting to emerge in the world. To be very aware and sensitive to that and then to be able to actualize that as it desires. Does that make sense?
The Key to Living An Active Life: Surrendering
W. Brian Arthur: Yes, absolutely, totally. I’m in unfamiliar territory here because my thoughts are about economics and some about business. I’m not sure I have thought much about spirituality, but I’ve certainly been through a lot. I’ve never tried to put them all together. I think in some strange sense the absolute key to living a very active life is to surrender; and what Buber says there is that it is not your own will, it’s a higher, deeper will. In some sense I think that one has to say, "Look, I’m here. I’m willing to do whatever is necessary for whatever reason. I’m here and give me the chance to do it and the means and I’m willing. The problem is I think that you can talk about inner knowing, but my instinct is to try to get ahead of that and say, "Therefore, here’s how you can train management." This is really your area more than mine. I haven’t thought much about that, and I’m not sure how one could do it or would do it. I can sense certain things. And I can say that I’ve seen many, many things that I could not explain by supposedly rational means.
At the Bottom of Science Is the Unknown
In all these sciences, there may be rivers of thought, and if you track each one back to its source, meaning the primitive concepts, you’re in the unknown. In physics, for example, you can track it all back to energy and really fundamental sub-particles like quarks and then further back. When you get back to those source concepts you’re in the unknown. We don’t know why there is energy. We don’t know what quarks are. We don’t know anything fundamental, and it’s a conceit that says science understands. It doesn’t. It starts with a few magical unknowns. It starts with the unknown and labels part of it and knots a few strings below that and then hangs onto those strings, but they’re not suspended from anything.
JJ: That’s so important.
W. Brian Arthur: It’s the same with chemistry. When you start to get into the nature of chemical bonds, you’re back to the same problem of energies and so on. It’s the same in economics.
I asked a friend of mine who is a philosopher why all I do these days is read philosophy. I can’t stand reading economics anymore, but I did twenty years ago. He said, well, if you drill deep enough into any subject, you get into philosophy. But philosophy is our way of coming to grips with the
39
40 unknown. Philosophy isn’t about the known. Philosophy is about the unknown. Philosophy is like scouts that we send out to explore new territory.
The point I want to make with all of this is we’ve constructed this edifice called Western science that happens to be good at a few things like CAD imaging, or guiding satellites or nuclear bombs, but it doesn’t mean we really understand the world at all. The only people who think we do are people who don’t understand science. People like Bohm or Einstein, who really do understand science, will tell you that there is a thin layer of what we do understand, but down below or above it we don’t know what we’re in.
Once you see that, then I don’t see why there should not be fields or areas of understanding that don’t have much to do with the cognizing process. This is something that I would never stand up and say. It is certainly my own belief. I’ve seen far too many phenomena that I can’t explain, and I don’t even know where to begin.
I don’t know where this would go with management. Images come into my mind of actually having people stop, calm down, sit, and let things emerge. Like if you’re forty-five and you have a mid-life crisis or something, and you go off to some island for two weeks. Maybe nothing happened there, but you may come back and say, "I need to change this and this. I’m not happy in my job," or something like that. I think that that is really good.
The current view is that the world is deterministic and mechanistic and that certain modalities can operate, problems, solution, subject, object, and so on. We tried, during the 20th century , to push those views into mathematics and physics and, for that matter, economics. Each place we pushed, it ran up against its limits. Certainly in physics, quantum theory, fundamental particle theory, in economics, the whole crashes upon itself similarly in philosophy. Wittgenstein tried to push it to its limits with the Tractatus. In his later philosophy, he had to let go of the idea that you could mechanize philosophy and make it into a logical consistent apparatus like a mechanical apparatus.
The Root of the Mechanistic View: The Catholic Church
Interestingly, all that came not from Descartes, but the Catholic Church. The mechanistic view came from the Church. In the early 1600s there were a lot of pantheists still in Europe, and Christianity was progressing. The Church thought it could do out the pantheists by saying the world was a mechanism.
There are corpuscles in the world, and those corpuscles (maybe we call them atoms) operate mechanistically. Everything is precise and there is one God operating that.
The Other View: An Organic Perspective
The pantheists’ point of view is that the world is organic. It all has its own spirit. Every stream or riverbed and tree and so on. They wanted to do away with this.
I think we’re currently going around one layer of a spiral. We have deeper understandings because of three or four centuries of science from Michelangelo, Galileo, and Descartes. But as we push up against the limits of the mechanistic view, we are finding that we’re coming back to the organic view that everything is unfolding, it’s all organic and it’s basically just interacting patterns. I don’t know what 40
41 consciousness is, I haven’t a clue, but it came to me one time in a meditation that consciousness is the universe being aware of itself, or pattern being aware of pattern. I really don’t know what that means, but we’re coming back to a view that is pattern oriented.
Complexity theory is the symptom, not a cause. Complexity theory is coming out of all the sciences and all the arts. We’re seeing that the world is structured in a formational view. It’s a view where biology displaces physics, where Darwin displaces Newton, where the computer displaces pencil-and- paper analysis. There are many movements here, but we’re coming into an organic view. Once we’re in an organic view, then the separations we made – subject/object, problem/solution, and so on – don’t make sense anymore.
When you do away with those distinctions, you’re in a completely different set of problems for management. The odd thing is that the more complicated and developed that technology gets, the less mechanistic it becomes, and the more organic. This is true of everything. The Internet is essentially very organic. It builds from what’s laid down already. It’s not easily describable. It’s not very homogenous, and it tends to reach out and just unfold.
GJ: It’s impossible to control.
W. Brian Arthur: Exactly. In the so-called real world, the economy itself is becoming more and more organic, and therefore the people who operate in the economy have to take that into account. The whole zeitgeist in itself is becoming less mechanistic and more organic. In turn, we’re much more conscious.
There is a continuum of problems, and some are very straight problems. Say I schedule your delivery trucks. It’s a nice problem. It’s well defined and so on. How you structure that will determine what happens there in the next five years. Now we’re moving into an area where management becomes less about scheduling fleets of delivery trucks and more about defining this whole area. A lot of what we were talking about has to do with different layers of what cognition means. There is no simple answer to that. Cognition might simply mean rummaging around in your attic for the frameworks and then saying, "I like this one. We’re really in a such and such situation." Or it could mean deep inner knowingness. I think that the East has an understanding that inner knowingness exists and that it’s worth something. But, again, this is a place I am totally familiar with and you are too, but it’s very hard. I’ve never thought about this for managers.
XI. The Business of Business Is Cognition
If you start to look at business as an entity, if the business of business is cognition rather than optimization, then you see business totally differently.
JJ: Right.
W. Brian Arthur: The more high tech a business gets, the more the business is a cognitive one. That has very different implications. For example, if I’m a New York bank and I want to get into digital banking and I know it’s going to be an appendage of what we’re doing now, but in ten years’ time there are going to be a half-dozen large digital banking networks, then how do we do it? It won’t be tough like it is now if you have to get into your bank account. You’ll be able to take $10,000 and drag and drop it into a mutual fund for the next three weeks. Now the problem is that there is layer after layer of 41
42 what to do, but the first thing I’d do is what Sony does and tell you to take your very best people and immerse them.
The people who win high-tech games are the people who cognize best – I wish I had a better word than "cognize" – the people who can frame it most accurately. If you can figure out what the game is rather than how to play it – it’s pretty obvious how to play it – then you’re going to do well.
If you think of the game as simply an extension of what went on before, you’re going to get shoved out. So if this is a very large bank, you can take a couple of senior people and send them to Seattle or California to immerse them for a couple of years. But the moment they come back they’re in for a very frustrating time. Suppose they come back and you say to them, "Fine, you’ve got the message and you’ve convinced me. I’m the chairman, I’m convinced, but you know it is contrary to everything I’ve done in my career, the board isn’t going to like it."
In a cognitive economy, most businesses are going to be small. They will have the ability to adapt faster; they’ll be more flexible. There isn’t a prayer of changing the large companies, but you can immediately set up a new one. I think that’s the way to move forward.
The Question
Let me put a question to you then, because I think this is an open question. You should mull this over. Every era of business has different outstanding problems. When production lines came along, the problems had to do with balancing production lines and getting people operating in an automatic way. At that point you had Taylor and people like that; you got time and motion type of management consulting and so on. In the Depression, the problem was how to shut down companies, and McKinsey was in that business. The ’50s was an era when how very large companies were structured made a huge difference. Strategic planning, a la McKinsey or Booz Allen, became very important. There might be another era now where you could say it’s all about mergers and acquisitions and putting together this and that, and some people do that very well. Now, what would a consulting company look like if it were in a cognitive economy?
This is more than teaching people to learn, it’s actually sense-making. It’s not just coming in like a McKinsey and saying, "The world has changed and we want to spend several months with you figuring out how things have changed and what businesses you ought to be in." That is done by every consulting company. This would go deeper. This would say, "Business after business comes to me saying they’re facing 90-degree turns, meaning that their whole business could go in a completely different direction. Not 180 degrees, not 30 degrees, but 90 degrees, and what do we do about it?" Usually this has to do with the digitalization of industry.
My response would not be to go set up this or that, but to get your best guys and put them out there where they will absorb. It’s not what people know that counts; it’s what they take for granted. What’s taken for granted around here is very different from what’s taken for granted in New Jersey. You need to put them in a more useful cognitive atmosphere, where they absorb the culture and learn to see what’s taken for granted.
I just want to put it as a question to you because I think that there is a business there. It’s a very worthwhile and valuable one: How do you operate in an area where the main problem of management 42
43 is becoming how to cognize and mentally structure a very changing environment, where the skills to be brought to bear on it are not trivial? They’re rather deeply understood, if they’re understood at all.
Social Embeddedness of Knowledge
As I said earlier, your friend and some friends in Japan say you could have all the deep insights in the world, but they would not know the kind of debts that would be needed to bring to bear the skills to set up companies like that. I could download the deep secrets of creating Stradivarius violins, but unless I knew where to get the wood and the glues and the people who understood all that, I could do nothing about it. This knowledge was embodied in people in the 1700s in a small town in Italy. It was not just that they understood how to build violins, but what was taken for granted about the various components. The way the wood took shape. The way it was aged. What degree of moisture you leave in it before you put the resins on it. All of this is embodied not in just one person, but in the whole surroundings of that culture. When that starts to get lost, it’s not sufficient to put it back. In that sense, cognition doesn’t just stop with one person’s understanding. It has a large infrastructure as to what your neighbors’ cognitions are, what is taken for granted in that culture, and then what is physically available in that culture to work with. I think that’s why regions often get ahead and stay ahead for centuries.
From that point of view, Silicon Valley is a set of cottage industries. It’s a set of little Stradivarius groups. They’re actually downstairs here. I can hear them, but I don’t understand half of what they’re saying. It’s all jargon. People in Silicon Valley all know exactly what each other means, and they know where to get what they need, they know how to put it together, and so on. From that point of view, high tech is a cottage craft, and yet how to put it all together and make it work is a cognitive challenge. The Stradivaris knew what they wanted to do. They were making violins. The fathers had made violins; they streamlined and perfected the work and passed it on through their families. Violin-making was going to endure. If you made violins here, you would have to hit the market just right, and then fifteen months later it would be something else. Four months ago, the bright thing here was eBay. eBay is history now. The country is starting to find out about eBay. eBay is over. They’ve all made their $200 billion or whatever. The concept is there, it’s past, and then you go on to the next big thing. This whole area operates in a series of big things, but what’s appropriate in each era is a different set of circumstances. So it’s a very different set of cognitive challenges.
Precognition
COS: In your Harvard Business Review article you use the term "precognition," and you say Bill Gates would be pretty good at precognition. What kind of cognition does it take to arrive at such a precognition? Wouldn’t that be the thing we talked about previously? This is, you have to access the level of knowing rather than of just conceptual knowledge? How do those relate to each other?
W. Brian Arthur: I think this has to do with conscious cognition. Having a very wide inventory of experience at an appropriate level in an industry is unbeatable. But if my inventory of experiences is in Pepsi-Cola and I’m brought in to Microsoft and then told, "Your challenge is to bring us into the Internet age," then my first question is, "What’s the Internet?"
If you have a long history at that level of deep knowledge then maybe you can do what good cooks do. "Well I can combine this with that. I know this works. That tastes good, but then if I add that bit of 43
44 ginger to offset that, it will be even better." If you can’t cook, or you haven’t used those ingredients before, you can’t do it. It’s that sort of knowledge.
There is a paradox, because if you are only inside the industry, you’re used to what’s being done already. So the idea is to be outside, to gather fresh recipes, fresh ingredients, learn new combinations, and then come back to your old position. But if you are in a pastry program and you’re asked to do lamb, you’re not going to get very far.
I find that most of the ways of doing digital business are 90 percent digital and 10 percent what the industry is about. For example, digital banking would be 10 percent finance and 90 percent software plus telecommunications. The people who are going to do well are going to come out of the software and telecommunications industry. They’ll buy the bank. Microsoft can take over Citibank and buy it eight times over, not vice versa. The opposite notion is, I’m a New York Bank and really I want to extend my services into digital banking. I’ll buy myself a few software programmers and we’ll do what we’re already doing, but we’ll be able to do it on the Internet. That’s not the way it’s going to work. The way it is going to work is people who have had experience at Amazon, Microsoft, and eBay are going to get together and say "Okay, you do the encryption, you do the clearing, you do the interface, let somebody else do such and such; oh, and we need a bit of finance. I know somebody in New York who has a bank. We’ll buy the bank." The bank will get absorbed and it will cease to be physical.
JJ: We feel like it has been a privilege to be with you.
W. Brian Arthur: I’ve been delighted. Thank you.
XII. Reflection
In this interview, W. Brian Arthur made three significant points: One, in order to understand today’s world economy, we need a different theoretical foundation of economic thought. On this point, Arthur is best known for his work on the economics of "increasing returns," which suggests a much more dynamic, fluid, and unfolding view of the economy. Two, what it takes to operate in this environment is a different kind of knowledge and knowing: a knowledge that does not stem from an abstract framework that we apply to or impose on a situation, but a knowing that emerges from the quietness of a deeper place. And three, what it takes to access this deeper source of knowing is to follow three steps: (1) total immersion: observe, observe, observe; (2) retreat and reflect: allow the inner knowing to emerge; (3) act in an instant: bring forth the new as it desires.
XIII. Bio
W. Brian Arthur is Citibank Professor at the Santa Fe Institute. From 1983 to 1996 he was Dean and Virginia Morrison Professor of Economics and Population Studies at Stanford University. He holds a Ph.D. from Berkeley in Operations Research, and has other degrees in economics, engineering and mathematics.
Arthur is best known for his work on positive feedbacks or increasing returns in the economy–what happens when products that gain market share find it easier to gain further market share–and their role 44
45 in locking markets in to the domination of one or two players. His work on increasing returns won him a Guggenheim Fellowship in 1987 and the Schumpeter Prize in Economics in 1990. It also won acceptance in Silicon Valley, where strategies based on increasing returns ideas now dominate high tech thinking. And it became the basis of the US Dept. of Justice vs. Microsoft case of the late 1990s. His papers on this topic were published in Increasing Returns and Path Dependence in the Economy, U. Mich. Press, 1994.
Arthur is also one of the pioneers of the new science of complexity–roughly speaking, the science of how patterns and structures self-organize from simple elements. His work here is detailed in Mitchell Waldrop’s 1992 book Complexity. His current interests are the economics of high technology; the "digital economy"; how business evolves in an era of high technology; and cognition in the economy. He is writing a book on high technology and the different economy it is bringing into being.
Arthur was the first director of the Economics Program at the Santa Fe Institute in New Mexico; and he currently serves on the Board of the Institute.
45
46
W. Brian Arthur is a professor at the Santa Fe Institute. Formerly a professor at Stanford University, he holds a Ph.D. from Berkeley, and has graduate degrees in economics, engineering, and mathematics. Arthur's main research interests lie in theorizing about the economics of the high technology sector. He is best known for his study of positive feedbacks, or increasing returns -- in particular their role in magnifying small, random events in the economy. He is the author of Increasing Returns and Path Dependence in the Economy. His ideas have attracted increased public attention with the recent antitrust investigation of Microsoft by the U.S. Department of Justice. Joel Klein -- Assistant Attorney General, and head of the antitrust division that is investigating Microsoft -- recently singled Arthur out as a theorist who has particularly influenced his thinking on high-technology markets. His work on increasing returns won him a Guggenheim Fellowship in 1987 and the Schumpeter Prize in Economics in 1990. He is a consultant to Citicorp and McKinsey and Co, and a Fellow with Coopers & Lybrand.
46
47
At the end of April, as the Microsoft case approached a new climax, Professor Arthur gave an interview with PreText magazine editor Dominic Gates. For the first time in a public forum he spoke extensively about his theories, his critics, and the Microsoft case. His Economic Theories:
Increasing Returns and Path Dependence U.S. Department of Justice versus Microsoft Answering Critics of his Theories Academic In-Fighting in the Media
47
48
http://www.pretext.com/may98/contents.htm
His Economic Theories: Increasing Returns and Path Dependence
Gates: What are increasing returns and how do they work?
Arthur: Increasing returns are a form of positive feedback: the tendency of anything that's ahead in market share to get farther ahead, or if something's falling behind in market share to get farther behind.
Gates: Is this a general mechanism in economics?
Arthur: Yes. It's a body of economics that's exactly parallel to diminishing returns in neoclassical economics. It's the twin concept. Say you've seen many, many movies in a month; sooner or later you've exhausted the ones that are matched for your taste; so you're running into things that are not so good. In industry, if you expand your business beyond a certain degree, it's always been held that you run into increasing costs or declining clientele, or increased hassles. You run into diminishing returns.
By contrast, very often in high tech as something gets farther ahead it gets more and more advantage. If many people are using Java, then people like me feel they have to install Java instead of, say, ActiveX. If a product gets further ahead it gets further advantage. If that's so, you say there are increasing returns.
The fact that there are increasing returns is wonderful news. If something gets better, as it's more used, this is great news; if something gets cheaper the more it is produced, that's wonderful. Diminishing returns made Carlyle call economics a dismal science. Increasing returns maybe makes economics a cheerful science.
Gates: There does seem to be a confusing proliferation of words to describe several closely related economic ideas here, so I'd like to define some terms. What is path dependence, and what are network effects?
Arthur: In high tech there are three very particular mechanisms that make for increasing returns. One is up-front costs. Because high tech products are extraordinarily complicated to design,
48
49 there are often very large up-front R&D costs; $200 or $300 million for Windows 95 for example. That's the first disk. And then the second and third disk out the door cost Microsoft only a few pennies.
Another effect is what I call customer groove-in. Sometimes it's just called learning effects. As you use something that's complicated, you come to know its foibles. And the more I type on the QWERTY typewriter keyboard, the better I get at that. It's harder for me to switch over to some alternative keyboard. The more an airline knows how to use, maintain, and fly Boeing aircraft, the more likely they are to order from Boeing. This seems to be very much a high tech thing, because it's hard to learn a high tech product. So the more I use Microsoft Word, which I do use, the more locked in I am or grooved-in to Microsoft.
[Then there are] network effects, sometimes called network externalities. In a nutshell, if nobody has a telephone, the telephone isn't much used; but if everybody has a telephone, or say e-mail, then it is much more useful. It's another form of positive feedback. If everybody is using Java, then more applications are going to be written for Java. It's the positive feedback that comes from a network of users.
Gates: And path dependence?
Arthur: Imagine there are increasing returns in the market with several products, all competing: Microsoft Money versus Quicken, or ActiveX versus Java. They're competing and then one of them gets ahead. If there's enough increasing returns there, as any one of them got ahead it would get further advantage.
When you have increasing returns, at the outset markets are unstable and lurch back and forth according to different small events, and then lock in to one of many possible outcomes. What locks in is a function of what happened in history. The outcome in increasing returns markets depends on small events along the way. The shorthand for that is "path dependence." Meaning that small events along the way decide the outcome.
The U.S. presidential primaries are very path dependent. Depending on who gets ahead, there are increasing returns. Candidates who look more presidential attract more money; if they have more money, they get more television time; with more television time, they're more likely to look like a presidential prospect. The outcome is decided by the pathway the whole process has taken through Iowa and New Hampshire.
Gates: Once you start down one path, whatever it is, it's hard to go back.
Arthur: Yes. It grooves-in a path. If rain falls on top of a sandy mountain, pretty soon it'll groove a pathway down the mountain and small events at the start will determine the topography and what rivers eventually form. It's important to note that the outcome is not completely determined by what's best. The outcome is partly determined by who gains what advantage when.
U.S. Department of Justice versus Microsoft
Gates: How did you become a player in the Justice Department case against Microsoft?
49
50
Arthur: In 1990 I published an article called "Positive Feedbacks in the Economy" in Scientific American, which was noticed by many people. One of them was Gary Reback [the lead lawyer representing several of Microsoft's corporate rivals, including Netscape and Sun]. When certain companies in Silicon Valley were filing a white paper against the acquisition of Intuit by Microsoft -- I think in the fall of 1995 -- Reback contacted me and asked me to take part in that. Actually, by then, I thought the case had a lot of merit. I was happy to be part of that.
And it seems further that that early paper of mine in Scientific American, and some of my other early papers, appear to have influenced people at the Justice Department -- notably Joel Klein -- and many others in academia and in the American legal system.
So when Microsoft became an issue, these sets of theories that I and other economists had been responsible for were what people started to reach for to make sense of those markets.
I would be pretty clear that these are theories of how those market operate; they're not theories of whether Microsoft is good or bad. Microsoft itself has used increasing returns theories to justify its position -- notably by saying, "Isn't it better to have a single standard?" Increasing returns theories are in no way against Microsoft. They have to do with how the markets operate. Both sides, the Department of Justice and the Microsoft lawyers, are obviously going to build their case upon how those markets operate.
Gates: Since that initial white paper, have you been actively involved in the Justice Department case against Microsoft? Do you anticipate a further role?
Arthur: I have been in conversations with the Department of Justice, but I am not directly involved. Once in a while I run into someone like Nathan Myhrvold; I'm not involved on the Microsoft side either. I describe myself at the moment as an observer: I'm extremely interested in the case, but by choice I've stayed on the sidelines simply because I wanted to write and think rather than be flitting back and forth to Washington.
Gates: So if you were called to give evidence, would you turn that down?
Arthur: Oh, I'd rather not make any statement on that. I think it would depend on where and when and so forth. Let me just not answer that. But you can certainly ask me my attitude to the case.
Gates: OK, let me ask that then. What do you think of the case, the Justice Department's case against Microsoft?
Arthur: First of all, I think that monopoly in high tech is not necessarily a bad thing. Monopolies are inevitable in high tech given that there are so many sources of increasing returns. But they are short-lived, temporary monopolies.
The important thing for the consumer in high technology is that innovation continues at a reasonable pace. [Because] you can achieve a product monopoly in high tech, if you're lucky you can go from being a graduate student one year to being worth half a billion dollars after an IPO three years later. This has acted as a very powerful incentive to innovate. Locking in the market becomes a prize for innovation. 50
51
I definitely think that the Department of Justice should not necessarily try to inhibit firms in the industry trying to lock-in this or that market. And by the way, the Department of Justice agrees with that.
What worries the Department of Justice, and also what would worry me, is if someone achieves a lock-in and then uses that unfairly in another market. The metaphor I've been using is that these markets are a little bit like the land rushes in the 1880s, in Oklahoma or Kansas. Everybody starts off with their horses and buggies behind the same line. [But in high tech] there's only one prize: you take 85 or 90 percent of the market.
Now imagine if you had won three or four of these races in succession and you parlayed your winnings into buying a Toyota Landcruiser instead of a horse and buggy. And just to make doubly sure, you hobble everybody else's horses at night. Well then I think that is cause for alarm.
To summarize my viewpoint, I think that the spotlight in high tech ought to be on two things that sometimes conflict: innovation and fairness. It's like the wild west in high tech at the moment. It's a wild and woolly and wonderful industry, like Dodge City or Tombstone in the 1880s, where anything goes and there's very little law. There's a huge amount of innovation and not that much regulation. The very idea that if you're lucky you can achieve a lock-in and get very seriously wealthy, that appears to have done wonders for innovation.
However, if these markets get very heavily locked in by a single firm for a long time, that appears to impede innovation. The fact that DOS had such a heavy lock on the operating system market for PCs meant that Microsoft didn't have a terribly strong incentive to innovate. It did indeed take them about 10 years to come out with a decent version of Windows.
The spotlight should be on innovation; achieving monopolies can be a prize for innovation. On the other hand, if one firm starts to monopolize everything, competitors can get scared out of the market and the dominant firm can have very little incentive to innovate. The right question to ask, in my opinion, is how is innovation affected?
And then the second question to ask is, what is competitively fair? It's not competitively fair to take over, say, 60 million users in Windows or 120 million or whatever it is, and then lever them onto your version of the Internet versus somebody else's. The Justice Department is right to be worried about that.
So yes, you can have a monopoly. But unfair use of the previous monopoly in a new market is not right. And again, I'd say that if you've won five races in a row, you can't just parlay that into winning the next five. It's healthier if everybody has a decent chance.
Gates: Well, if Microsoft is trying to do as you're suggesting, do you see Bill Gates as commanding a new evil empire?
Arthur: [laughs] No, I don't think it's an evil empire. There is healthy competition and then there is unfair competition. I think the evidence is steadily mounting that Microsoft has over the years crossed the line many, many times.
51
52
Gates: So do you think high tech needs to be regulated by the government?
Arthur: I think the government needs to do what is necessary to maintain innovation in the market and to maintain fairness. I think it's an extraordinarily tricky business to try to regulate in high tech markets, because they're moving and changing so rapidly.
I would not say that the government should want to regulate a specific market so as to try to influence outcomes or pick winners. On the other hand, I do think we need some rules of the game, just as there are rules for rugby or ice hockey.
I actually think that that's what's going to happen over the next five to ten years, especially because this Microsoft case is likely by precedent to set down a lot of the rules for high tech.
Gates: You mentioned Assistant Attorney General Joel Klein. In a New Yorker article, John Cassidy quoted Klein as singling you out as an economist who had greatly influenced his thinking about the way in which high tech markets operate. Cassidy went on to quote Klein as saying that "surgically applied [government] intervention" is desirable.
Do you have faith in the surgical application of government power?
Arthur: [laughs] Well, like most economists, I believe in free markets. I'm not alarmed that increasing returns spell doom to the capitalist system. However, I do think that there should be rules of fairness. What the government needs to do is to set down rules of the game for fairness and rules that preserve innovation.
I'm not sure that even if it wanted to the government could go into particular high tech markets and regulate how the outcome should be. And in fact, I'm not sure that's what their intention is either. As far as I can see they are not going to rule out the notion of monopolies.
But certainly there should be rules against what you might call night maneuvers -- maneuvering or scheming in a pernicious way and not competing fairly. And sad to say, there has been unfair competition in the software industry and in high tech in general, and I do think that rules need to be laid down as to what is fair. And enforced.
And again, more generally, we're leaving a period where the dominant pattern in the economy is bulk processing of steel, and automobiles, and cement, and heavy chemicals, and lumber, and so on. We're entering the era of high precision and fast moving high technology. All the rules of the game are written for this bulk processing era. The rules need to be rewritten for high tech. I would imagine they're going to be slightly different.
But as an economist, I would certainly lean towards minimal government intervention; and I lean toward allowing monopolies in high tech -- at least the smaller, temporary kind -- because those are the incentives to get rich and to innovate.
And then, if there are monopolies that are simply not playing fair, I would go after them. No question.
52
53
Gates: Cassidy's New Yorker article, which is very flattering to you, also had an interesting quote from Supreme Court Justice Learned Hand, who wrote "The successful competitor, having been urged to compete, must not be turned upon when he wins." That's a persuasive sound bite against too active an antitrust stance. Do you think that applies to Microsoft?
Arthur: No, I don't think so. I don't think that Microsoft is being taken to court because it's a winner. In fact, a large proportion of the American public and politicians revere Microsoft and Bill Gates simply because they've been winners.
Microsoft's being taken to court because it's felt that it's making unfair use of its monopolies. I think it's up to the government to prove it; my hunch is that there will be a lot of evidence in favor of that proposition.
I think that the people at the Department of Justice, the ones I've talked to, are certainly sophisticated about that. They're not going after Microsoft because it's Microsoft, or because it's highly visible. They don't want to see precedent set in high tech that lock-ins can be levered from one market into another. Or that the smaller, weaker competitors can be hobbled.
Gates: Another aspect of antitrust actions is the reality of legal procedures. The famous example is the IBM antitrust investigation, which took years, sapped corporate energy and money, and went nowhere. That raises the question of whether these cases are a feeding frenzy for lawyers and counterproductive in many ways.
Arthur: Yes, this is an element sometimes in these very large cases. Any legal system can provide nutrition for lawyers for many, many years. Cases like this should be brought reluctantly. But in the case of IBM, or the Bell phone system, or Microsoft, the public does have a right to find out whether their productive resources are being used efficiently and fairly.
Gates: Whose productive resources?
Arthur: The American public's resources. The productive resources of the country in the hands of Bill Gates or IBM. The public has a right to know from time to time if these monopolies are behaving efficiently and fairly. And from time to time large monopolies are going to get dragged into court to be probed. But, the answer to the IBM question is, if someone's brought into court and found innocent of murder, does that mean we shouldn't have any more murder trials?
Gates: Is antitrust intervention what moderates and makes acceptable the operation of market capitalism?
Arthur: No. Not at all. There are laws and institutions and stock markets, and then there's a tiny hair on the tail of the dog that is antitrust policy. It's one among many institutional backings to capitalism, that's all. And for the most part, it's never evoked. Competition goes on every day and there's very, very little intervention on an antitrust basis. But once in a while, large companies are called to Washington to justify what they've done. That's as it should be.
53
54
Answering Critics of His Theories
Gates: Is the theory of increasing returns still controversial?
Arthur: Absolutely not. This is now completely taken for granted in Silicon Valley. I don't have to go around California telling the Marc Andreesens of the world or the Andy Groves of the world that there are increasing returns. Intuitively, the smart people in high tech knew this all along.
Gates: Do you think Bill Gates realizes this?
Arthur: Absolutely, and has done for a long, long time, independent probably of any academic theories.
I don't know anyone who would describe themselves as a market capitalist, as a typical business person, who would find increasing returns threatening. On the contrary, what we're finding is that these are a body of theories that resonate very deeply with their own intuitions. Folks at Sun Microsystems, or other places, are using these theories. Sun used my theories to launch Java. In return they gave me a high end Sun workstation.
And all the academic battles about increasing returns were over around 1990. That's when the controversy stopped over whether it was correct, and the controversy started as to who had thought of it first.
Gates: So you don't see these ideas in opposition to classical economic theories, the Chicago school?
Arthur: Not at all.
However, some people who are great proponents of Chicago neoclassical economics seem to get uptight every so often in the opinion pages of the Wall Street Journal. The source of the problem is that if I say that markets can lock in to one product or one company, not necessarily the best, then that's taken as a threat to the whole ideology of capitalism.
The only controversies are ideological ones. I think it's inevitable that any important theory, or any new theory of any importance, does have a trail of flat-earthers behind it, a trail of creationists; people who won't get it and don't get it, for one or other ideological reason.
So there is a rearguard battle being fought between academic economists and, how shall I put it, capitalist ideologues. Two or three years ago the Economist magazine, which can be quite ideological, seemed to be negative towards these ideas. Now it is no longer denying any of this.
Gates: And the Wall Street Journal?
Arthur: Well, the Wall Street Journal itself has story after story of increasing returns and the dynamics of these market. Only the opinion page of the Wall Street Journal is a lagging indicator of economic thinking.
54
55
Gates: So these theories are no threat to capitalism?
Arthur: On the contrary. Markets operate according to diminishing or increasing returns. Those are just like the laws of physics. Markets operate the way they do. Capitalism is a structure that's built on top of those markets, and it seems to me that standard capitalism of the sort that we have now does very well indeed under increasing returns.
But it does tend to make [the more] highly open capitalistic markets, as there are in high tech, seem to be somewhat more unstable. People in high tech know that you can lock-in a market almost before it starts. I don't think that is a threat to capitalism, but it makes for a less leisurely capitalism. And it makes for maybe more intense competition.
Gates: Let's talk about some of your critics. Stan Leibowitz and Stephen Margolis have written a critique, which was published on the Web site of the Libertarian Cato Institute and was the subject of a recent story in the Wall Street Journal. They claim to debunk the historical basis of path dependence theory, specifically the famous QWERTY story [that the familiar QWERTY arrangement of the keys on a typewriter was deliberately designed in the 19th century to slow typists down, because early manual typewriters tended to jam. Once typewriter manufacturers were locked into QWERTY, an alternative design that allowed faster typing failed to supplant it.] Leibowitz and Margolis call this story a "fable" and the Wall Street Journal refers to it as an "urban legend."
Arthur: It is perfectly demonstrable that we are indeed locked into a single QWERTY keyboard. There are legions of examples of lock-in. I'm not sure even Margolis or Leibowitz would deny that.
Gates: Right. But what they were saying was that it wasn't an inferior technology that locked in, and that the historical story which claimed it to be so was simply not true.
Arthur: It's absurd to think that any theories of increasing returns hinge upon whether QWERTY is better or worse. That is nonsense.
If you shine the appropriate light on it, you could demonstrate that under certain circumstances something that locked in -- like QWERTY -- wasn't so bad after all. I don't know anybody who is saying QWERTY is wonderful, but it's not clear to me that QWERTY is that great.
Take another example in the Wall Street Journal article: Microsoft DOS. I know of no independent computer scientist who thinks that DOS was a wonderful operating system, even when it was produced, though you can find ingenious ways to show that it was in some strange sense superior.
Gates: Well, they claimed that DOS beat out Apple because it was cheaper.
Arthur: One can take anything that locks in and at the time it locks in, normally it's better; that's why people are buying it. It's more convenient, or it's out there, or it's what you run across. But the point is that there could have been something else that might have locked in that, in the long run, may well have been better.
55
56
Not so long after DOS came out, the Macintosh [operating system] was demonstrably better. I think that Microsoft itself has acknowledged that fact by designing Windows to look just like it. And if people are saying, "Yes, but DOS was cheaper", well, think of all the wasted hours trying to use the damned thing. In computer science circles, DOS was a joke.
As far as I can see, the Leibowitz and Margolis arguments are ideological arguments for the far right. I don't see that there is a debate on increasing returns. You can have a debate as to whether what locked in might, under certain lights you shine on it, actually be better than what was locked out. You can make a case that gasoline engines are better than any alternative could have been. But frankly I don't know how to settle that, because you're talking about what might have been versus what is.
Gates: These theories are often discussed via particular examples or counter examples. One that you have cited is VHS versus Beta Max. Leibowitz and Margolis say that there was a good reason why VHS won: VHS tapes could record longer. And so, there was a reason why it locked in; it's not an example of path dependence.
Arthur: It is an example of path dependence.
The question of whether the product at the start was better or worse is moot. Yes, people may adopt VHS because it has a longer recording time. But the point of increasing returns is that if it gets ahead it locks in. Not what is better or what is worse. That's only a point for ideologues and the back pages of the Wall Street Journal.
Gates: But isn't an important part of your contribution your pointing out that things that get locked in aren't necessarily the best? It's not just to demonstrate lock-in, but to demonstrate lock-in of something that wasn't good for consumers.
Arthur: Well, again, you only get excited about that if you belong to the right wing of American ideology.
This notion that the market is always wonderful and perfect is a right-wing ideological idea. People don't expect that all the friends they have are the most optimal friends. People get married; sometimes it's wonderful and sometimes it isn't. Lock-ins occur; sometimes for the best, sometimes not.
The theory doesn't say that what locks in has to be inferior. The theory says that it's not necessarily superior.
Gates: That same Wall Street Journal article concluded that there is "an emerging consensus . . . that the path dependence school has yet to come up with the smoking gun it needs to show the market- place locked into a manifestly inferior technology."
Do you have a smoking gun for increasing returns?
Arthur: I find I'm puzzled by all of this because it's a bit like debating evolution with creationists: "But if you believe in evolution, the inference is that angels must have evolved their wings, and that would upset all of theology." For me it's moot. The onus isn't on me or anyone else, to show that we're locked in to any inferior thing. The onus is on the opinion page of the Wall Street 56
57
Journal and the libertarians to show that all things that we're using in the economy are not just the best they could have been at the time, but are the best that could possibly have emerged. Nobody in computer science believes that about DOS. As for the QWERTY keyboard, if Margolis and Liebowitz can prove it's the best, my hat is off to them.
Gates: Let me throw at you some more of these free enterprise think tank critiques of your theories. Clyde Wayne Crews went beyond saying that lock-in to inferior goods was a myth; he claims that lock-in is a myth. The examples he cited were: color TVs did supersede black and white; CDs did replace vinyl records. In another piece, Robert Levy of the Cato Institute, added a couple more examples: Word Perfect once looked unassailable as a word processing product; Lotus 123 once had no competition in spreadsheets. All of those actually failed to lock in and exclude the competition.
Arthur: Not at all. They all locked in, but then the next wave of technology took over. We were indeed locked in.
The fact is, technology comes in waves. No one I know who talks of increasing returns says that lock-in is forever. We are locked in to English, temporarily. In 500 years time it'll be a different language. Three-hundred years ago people were locked into Latin as the international means of discourse. No one said a lock-in is forever. In fact, it's taken for granted in high tech that lock-ins typically last anywhere between a year or two and five years.
Let me give you a very specific example here again. Netscape, as you know, had a heavy lock- in in the browser market. And it wasn't dislodged by means of a new wave of technology: no new software product came along to supplant the browser; instead it was steamrollered aside by the Microsoft juggernaut, Internet Explorer.
Gates: But it hasn't exactly been steamrollered out of the way. It still actually has a bigger share of the market than Internet Explorer.
Arthur: Well, you can certainly claim that its lock in isn't as heavy as it was two years ago. I'm just saying that a lock-in is only good until the next wave of technology, until the game changes. And even if the game doesn't change -- it didn't with the browser market -- if you have enough guns, you can dislodge the lock-in.
Gates: Isn't lock-in just another word for standardization? Britain and the U.S. drive on different sides of the road. Wouldn't it be better if they both drove on the same side and you only had to make one kind of car? Similarly, the European Union has a single cell phone standard and the United States has three incompatible technologies.
What's wrong with standardization?
Arthur: Increasing returns are about the dynamics of markets. If a market locks in to something, it's not necessarily the best; on the other hand, as you were saying, there may well be advantages to locking in to a single standard. So any theories of increasing returns aren't necessarily pro- or anti-Microsoft. Under increasing returns, you can lock into a single standard and that might have enormous advantages.
57
58
Judging the pros and cons of increasing returns markets is case specific. Let me give you one example. If a market, say, software, locked into a single standard, say, Microsoft Windows/Explorer, you could argue that there's some advantage to that. It would be like having a single railway gauge all the way from Calais to Moscow a hundred years ago, so you didn't have to change trains at each border.
So my answer is yes, there are many advantages to increasing returns, and certainly one of them is that we can use a certain standard. Basically the entire Internet is the result of a telecommunications/computing standard: TCP/IP. The existence of that standard made the World Wide Web possible. So yes, there are advantages in standardization.
Increasing returns are in a particular industry. They're either present or they're not. I want to get my point very clear on this. Increasing returns have to do with how markets work. Whether that is good or bad is somewhat case specific.
Academic In-Fighting in the Media
Gates: [MIT economist] Paul Krugman, is in some sense also a path dependence theorist. One might expect him to be on your side. What did you make of his attack on you a few months ago in Slate?
Arthur: I think it says that these theories of increasing returns are well accepted, and now the fights are not over who's right or wrong, but who said what first.
I don't want to leave the impression that all of this came out of nowhere in the 1980s; it had quite an ancestry. Economists have bandied about the ideas of increasing returns for over 100 years. Some of the really great economists, such as Alfred Marshall in 1891, had asked the question: instead of diminishing returns, what if firms gained an advantage the more market they took? I didn't invent increasing returns any more than Paul Krugman did.
But until my work in the early '80s, increasing returns was a static subject. Economists were aware that different outcomes were possible, with one or other product locking-in to market dominance. What was missing was a theory of the dynamic process by which market lock-ins take place. I supplied that theory, and did it rigorously, using modern, nonlinear probability theory. I have a series of articles and a book about that. This brought a greater awareness of dynamic labels and terms such as path dependence and lock-in. As far as I know, I was also the first person to claim that this had a huge area of application in high technology.
As I see it, Krugman's attack is his attempt to rewrite economic history, to stake claims and write in a very minor role for me. My attitude is, let history be the judge. What I've said is in writing, and the dates are marked; and what he said is in writing.
Krugman is a good economic theorist, and Paul is somebody I've been talking with for many years. It's a shame now that there are fights over precedence and who said what. Paul and his friends have done excellent work on international trade theory under increasing returns, and some newer work on regional economics under increasing returns. But they did not do anything about the dynamics of lock-in or small events being magnified. 58
59
I think Paul should just keep his hair on and do what he does. He's very good at it.
59
60
The Force of an Idea o
John Cassidy
New Yorker 12 Jan 98 32
Until recently Bill Gates and Microsoft seemed unstoppable. Now its a different story. What Happened?
• Judge Declares Microsoft a Monopoly Nov 1999
In a way, Bill Gates's current troubles with the justice Department grew out of an economics seminar that took place thirteen years ago, in December of 1984, at Harvard's John E Kennedy School of Government. The guest speaker was Brian Arthur, a little-known Stanford economist who was having difficulty getting his articles published in professional journals.
The paper Arthur now read "Competing Technologies and Lock-in by Historical Small Events: The Dynamics of Choice Under Increasing Returns” drew a strong, and largely hostile, response. One Harvard economist, Richard Zeckhauser, stood up afterward and said, "If you are right, capitalism can't work."
A few months later, when Arthur read the same paper to a gathering in Moscow, an equally eminent Russian economist declaimed, "Your argument cannot be true!" Such was Arthur's challenge to economic orthodoxy that it would be another five years before he succeeded in getting this paper published.
Since the Second World War, economists have spent most of their time articulating the reasons that free markets work so well, and arguing that government intervention in the economy is usually not only unnecessary but often downright harmful. Their argument, which harks back to Adam Smith, was rigorously developed in the "general equilibrium' models of the nineteen' fifties and vociferously promoted by the Chicago School of the sixties and seventies, whose members included Milton Friedman, Gary Becker, and Ronald Coase, Nobel laureates aH.
The Chicago doctrine was translated into the policy arena by Robert Bork, whose 1978 book, "The Antitrust Paradox," became an important intellectual prop for the generation of conservative jurists who were appointed to the federal courts by Richard Nixon and Ronald Reagan. Arthur's paper argued that the underlying assumptions of the Chicago School simply do@t apply to large parts of the economy, especially the hightechnology and communications industries. In these fast-growing sectors, Arthur said, there is no guarantee that the market, left to its own devices, will select the best products and maximize benefits to the consumer. Instead, he maintained, inferior products can beat out superior products merely because of happenstance-by being first to the store shelf, say - and they can remain in a dominant position for a long time. Small events, such as a misleading marketing campaign, can be magnified into big changes in sates. And some firms are likely to establish, through predatory tactics or
60
61 mere luck, lucrative and lasting monopoees, which stifle the very competition that freemarket advocates swear by.
In those instances, government intervention may be needed to restore competition. When I met Arthur re cently, in Washington at Ralph Nader's two-day conference on Microsoft, he turned out to be a gray haired native of Belfast who speaks softly and views the real world with a detached sense of amuse ment. "I was saying all this during the Cold War, so ideology got in the way," he said, smiling broadly. 'I spent about ten years in the wilderness." He did'nt seem bitter about this fate, which was probably inevitable, given the revolutionary nature of his theory. "It stands a great deal of standard economics on its head," he said with another smile. Arthur eventually emerged from the wilderness.
His arguments couldn't be dismissed as "informal or anectdotal" the customary derisive tags for heterodox economic ideas - because they were expressed in exactly the sort of dense mathematics favored by the editors of the American Economic Review. Arthur had taken degrees in electrical engineering and mathematics before turning to the dismal science, and he still refers to himself as an "applied mathematician" rather than as an economist. He is now a professor at the Sante Fe Institute, where physicists, computer scientists, and economists are applying the nascent science of complexity which examines how fairly simple dynamic systems can produce incredibly complicated outcomes - to subjects as diverse as geography, inequality, and finance.
Gradually, a number of economists began to take Arthur's conclusions seriously. Some of them- including Garth Saloner, of Stanford, Joseph Fatten, of Berkeley, and Steven Salop, of Georgetown- went on to develop game-theory models that demonstrated how powerfid firms could exploit the peculiar nature of high-tech markets to the disadvantage of their opponents. Finally, even figures outside the economics profession started to take notice. "At first, people said, 'Your theory may be theoreticafly valid, but there's no actual evidence of it in the economy,' " Arthur recalled. "I thought about that, and said, 'No, no, no. The whole high-tech sector operates in this way.'When I started to say that, I discovered it had a lot of resonance in Silicon Valley. People I talked to there just nodded wisely, grinned, and said,'This is how we see it, too, but we've never seen it written down and formalized.' "
One of the people who picked up on Arthur's work was Gary Reback, a Palo Alto antitrust lawyer who represents several of Microsoft's rivals. During the nineteen-eighties, Reback came to despair of the Justice Department, the Federal Trade Commission, and the courts, all of which he believed were in thrall to Chicago School economics. In 1995, Reback asked Arthur and Saloner to contribute to a white paper that he was drawing up for the Justice Department to protest Microsoft's abortive takeover bid for Intuit, a company that manufactures software for personal finance.
Ever since 1981, when Microsoft provided the MS-DOS operating system for the first I.B.M. personal computer, the firm's rivals have claimed that it uses anti-competitive tactics, such as issuing restrictive contracts to its customers, announcing software products that don't yet exist, tying products to each other in such a way that customers have to buy both of them, and deliberately making its products incompatible with those of its rivals. Rebacles white paper reviewed some of these charges, related them to recent academic work on the subject, and concluded with a stark warning about the future of the on-line information world:
"The markets today consist almost entirely of American competitors. But without government intervention, Microsoft will in short order crush this competition." Gates has always denied acting in a predatory manner, and during the nineteen-eighties his rivals' complaints evoked htde sympathy in 61
62
Washington. Finally, in 1990, the F.T.C. decided to investigate Microsoft, but three years later four F.TC. commissioners deadlocked on whether to bring legal action. The Justice Department belatedly took up the case, and in July of 1994 Microsoft agreed to a consent decree with the government in which the firm pledged to stop charging computer manufacturers, such as Compaq and Hewlett- Packard, a license fee based on the total number of personal computers they shipped, regardless of whether those computers contained Microsoft software. (In effect, the contracts forced computer manufacturers who wanted to install non-Microsoft operating systems, such as DR-DOS or I.B.M.'s OS/2, to pay two licensing fees-one to Microsoft and one to the other company.) Microsoft didn't admit to any wrongdoing, and Gates said publicly that he regarded the consent decree as a minor setback that would have little impact on how he conducted his business.
Even now, with about ninety per cent of the market for personal-computer operating systems, Microsoft denies being a monopolist. "We have never admitted to it, nor has any court ever found that to be the case," William Neukom, Microsoft's top lawyer, told me. "In this industry, we do not have the sort of monopoly power where one can reduce the supply and increase the price to the point of extracting monopoly rent. It's the reverse. People are getting better technology at lower prices." In October, the Justice Department asserted that Microsoft, by forcing computer manufacturers to include its Internet Explorer (a piece of software used for viewing-and retrieving information from-the World Wide Web) with every copy of its Windows 95 operating system, had violated a term of the 1994 consent decree which barred the company from tying any "other product" to the purchase of its operating system. Microsoft vigorously opI posed the government's argument, citing another clause in the consent decree that expressly allowed it to develop "integrated products," and a bitter legal and public-relations battle ensued.
Last month, Judge Thomas Penfield Jackson, of the United States District Court in Washington, surprised many observers by issuing a preliminary injunction in the government's favor. In ordering Microsoft to stop bundling Internet Explorer with Windows 95, the judge did't say the firm had violated the consent decree, the language of which he pronounced ambiguous, but he declared that Microsoft was indeed a monopolist, and suggested that its behavior may have violated the antitrust laws. "The probability that Microsoft will not only continue to reinforce its operating system monopoly by its licensing practices, but might also acquire another monopoly in the Internet browser market, is simply too great to tolerate indefinitely until the issue is finally resolved," Judge Jackson wrote. Microsoft appealed his ruling, arguing that it was virtually impossible to remove Internet Explorer from Windows 95.
The case rapidly turned into an imbroglio, with the Justice Department accusing the company of failing to obey the court's order, Microsoft accusing the government of being "poorly informed," and Judge Jackson using a laptop in the courtroom to demonstrate how easy it would be to remove the Internet Explorer browser from Windows 95. The legal skirmishing is set to resume next week, but the ultimate outcome of the case probably won't be known for several months. Judge Jackson has appointed a Harvard law professor, Lawrence Lessig, to report back to him on the technical issues involved in the dispute before he makes a final ruling, and he has also asked both sides to provide fiirther evidence. (Microsoft is trying to get Lessig removed, claiming that he has been given too broad a mandate.)
As the case proceeds, the Justice Department will be relying on just the sort of intellectual arguments that economists like Brian Arthur have been making. "In these kinds of markets, it is just not right that leaving it to the market is always going to get an efficient outcome," Daniel Rubinfeld, a 62
63 former Berkeley professor who is now the chief economist in the Department of Justice's antitrust division, told me recently. "There is stin an honest debate about exactly what role government ought to play, and people are going to differ, but there are very few economists I have talked to who would argue that leaving it to the market is always the best solution. We are just not in that world anymore." Joel Klein, an intense, sallow fifty-one-year old Assistant Attorney General who heads the antitrust division, shares that view. Klein is a lawyer (before moving to jus tice he was a deputy White House counsel), but he studied economics at Columbia, and he credits Arthur in particular with influencing his thinking on how high-technology markets operate. "In the nineteen eighties, the view was that markets worked just fine, and the government basically ought to stay out-that the cost of government intervention out weighs its benefits," he told me. The "new synthesis," he went on to say, is "that markets do@t always self-correct, and that surgically applied intervention aimed at protecting consumer choice and preventing the abuse of market power is desirable."
In the Chicago world, competition doesn't need guarding by the Justice Department or anybody else. If a firm makes monopoly profits, that fact will attract new entrants into the industry and spur existing competitors to innovate. Before long, the firm's monopoly will be broken, and till-scale competition will be restored. It is an attractive vision, and one that Gates exploits in his public speeches. "In the world of technology, nobody has a guaranteed position," he said at Microsofts annual shareholder meeting last November. "New ideas like browsers and Java operating system, or software built on artificialintelligence technolo@if that's done very well and we don't do something that's even better, our leading position could be eroded quite rapidly." If the new economic theories are correct, Gates's fortune, which is made up of about thirty-seven billion dollars in Microsoft stock, is a great deal more secure than he claims. "There has already been a lot of technical innovation by other people," notes Steven Salop, the Georgetown economist, who has advised the Justice Department on the Microsoft case. "Microsoft wasn't first with the Web browser, but nothing has dislodged it. . "
According to Salop, Microsoft is a "powerfid monopoly' that is unlikely to be unseated in the near fiiture, even by "competitors who have better products." Microsoft's power comes from its ability to exploit what economists can 14 network externalities." (Arthur uses the phrase "increasing returns," but he is takng about the same thing.) In plain English, "network externalities" means that the value of a product increases along with the number of other people who are already using it. This is not generally true-few people care how many others are buying the same brand of soap or cornflakes-but it usually ap plies to high-tech goods, for two rea sons: they have to be compatible with one another (a Betamax videocassette player is of no use these days, because it can't play VHS cassettes), and they are often linked to a network, in which case the more people there are on the net work the more valuable the product be comes. (A telephone is worthless if you're the only person who owns one.) In a business with network external ities, such as the market for personal computer operating systems, firms that control the industry standards and boast a large installed base of products have an enormous advantage over their rivals. Most computer buyers know little about the operating system, which is usually pre- installed by computer man ufacturers, but they care greatly about the availability of a large number of ap plicafions, such as games, word proces sors, and spreadsheets, to run on top of the operating system.
Writing these ap plications is a time-consuming and costly process, so independent software developers tend to design them only for operating systems that have a large in stalled base. The result is a "positive feedbacle' process: successfiil operating systems tend to have more applications written for them, which further strengthens their market position. This, in turn, encourages more developers to write programs, which attracts more cus tomers, and so on. Eventually, the entire market tips into the
63
64 hands of one firm, and that company's technology "locks in." Competitors who arrive on the scene after a market has locked in face a tough task, even if their products are good ones.
Digital Research's DR-DOS, which was launched in the late eighties, was in some ways more advanced than the competing version of Microsoft's MS-DOS, but it failed to acquire a significant market share and ultimately all but disappeared. During the early nineties, I.B.M. spent about two billion dollars developing and marketing OS/2, a sophisticated rival to Windows, but it, too, flopped, despite laudatory reviews from independent experts. "There is no presumption in markets with increasing returns that superior technology wins," Garth Saloner said recently. "An inferior technology that gets in first and is supported by network externalities may be able to hold its place even against superior technologies that come later." Many economists point to the VHS videocassette as an example of an inferior technology that locked in and defeated a superior alternative. (VHS was in some ways technically inferior to Betamax, but VHS was able to offer a larger selection of movies.) Arthur reckons that Microsoft's operating system is another. "People often ask me to give an example of something inferior that locked in," he said. "I say, look at MSDOS. Here is something totally crummy that locked in for ten years." Arthur's opinion of MS-DOS was shared by many computer scientists, who argued that it was a primitive and unwieldy operating system, but with the development of Windows Microsoft has greatly improved the quality of its products over the years, and few customers are complaining. At the end of 1997, it was estimated that Microsoft had a total installed base of about a hundred and fifty million personal computers.
The presence of significant network externalities creates powerful incentives for a dominant firm to try to manipulate the market. One way for it to do so is to make products incompafible, or partially incompatible, with offerings from rival venders, in which case customers will tend to stick with products that they know are idly compatible and reliable. Microsoft has been accused of adopting this tactic on numerous occasions. A few years ago, Microsoft told customers that DR-DOS would not work with Windows 3.1, which was then being tested, and those who tried got an error message on their screens. (DR-DOS designers maintained that there was no reason the two programs should't work together.) More recently, Microsoft has made arrangements to sponsor sites that are inaccessible to viewers who are using a Navigator browser from Netscape.
A dominant firm can also sabotage its rivals' products by announcing that it is about to come out with a simdar product, even if the actual launch date is months, or even years, away. The mere potential existence of such "vaporware," as it has come to be called, may well persuade buyers to wait for the dominant firm's offering instead of switching to another supplier, thus confirming the dominant firm's position. Garth Saloner and Joseph Farrell published a formal model in the American Economic Review several years ago showing how this could happen. "Especially when targetted against a fledgling technology, the pre-announcement may wen be anticompetitive," they wrote. Many software companies make strategic use of product pre-announcements, but Microsoft is the company that is most notorious for it. During the eighties, so many of the firm's products failed to materialize on titne (the original Windows was the most famous example) that Gates became known in Silicon Valley as the Viscount of Vapor. Perhaps the most important way a dominant firm can exploit its position, however, is by using its monopoly in one market to bludgeon its way into another.
This is particularly significant in light of the network externalities in the computer industry, since new markets are developing there all the time. "I think that high-tech markets are a bit like the land rushes that the United States used to have in the eighteen-eighties in places like Kansas and Oklahoma," Arthur explains. "Everybody lines up behind the starting line, they race their horses and 64
65 buggies, and if they win they get to stake out their hundred and sixty acres." The problems come, he added, if the victor in one race "parlays his winnings into a Toyota Landcruiser" to use in the next race, or if he "hobbles all the other horses at night." Joel Klein, at the Justice Department, uses a similar figure of speech, comparing the computer industry with a series of sprints, in which the government's role is to make sure that the contestants line up squarely on the starting line. "Our view is that a monopoly fairly acquired is not unlawful," he told me. "Our concern is with the use of monopoly power, once acquired, to protect or extend that monopoly-in this case, the use of monopoly power in operating systems to undem-line competition among browsers." It may seem odd for the Justice Department to be so concerned about the fate of a product that is effectively given away to many consumers by both Microsoft and Netscape, but the government believes that there is a lot more at stake in the battle between Internet Explorer and Navigator than the market for Web browsers. "Microsoft's own documents show that what Bill Gates cared about from Day One was that browsers could go after the opfrating system," Klein told me. "We have quote after quote from their executives saying, basically, 'We are not worried simply about the browsers, we are worried about control of the desktop, and therefore we have got to win the browser war.'That's the story."
Court documents filed by the government provide support for Klein's argument. In one such document, an internal memo dated May 26, 1995, Gates wrote, "The Internet is the most important single development to come along since the IBM PC was introduced in 1981. It has enough users that it is benefitting from the positive feedback loop of the more users it gets, the more content it gets, and the more content it gets, the more users it gets." Furthermore, Gates continued, the Internet presented a major threat to Microsoft because the firm's rivals, such as Sun Microsystems and Netscape, were trying to exploit and "commoditize the underlying operating system," which is the product that Microsoft's success is based upon. Since Gates wrote that memo, the potential threat to Microsoft has increased sharply. I.B.M. and many other companies are developing software applications, such as word processors and video games, that use Sun Microsystem's Java programming language to run on Web browsers, without any need for an underlying operating system, such as Windows 95. "Netscape/java is using the browser to create a 'virtual operating system,"' Paul Maritz, a senior Microsoft executive, warned in another document obtained by the Justice Department. He also wondered whether Windows would become "devalued" or "eventually replaceable." Beginning in mid- 1995, Microsoft's response to this challenge was to start bundling Internet Explorer with Windows 95, so that anybody who bought a computer with a Microsoft operating system also received a Microsoft Web browser, but Judge Jackson has outlawed this tactic, at least temporarily. In stating that Microsoft's right to integrate new products into Windows "stops at least at the point at which it would violate established antitrust law," he drew attention to a series of statutes that many people in Silicon Valley had written off as outdated and ineffectual.
The antitrust laws, as their name suggests, were introduced in response to the commercial combines that dominated the markets for sugar, lumber, beer, and numerous other goods at the turn of the last century. (The most famous of the "trusts" were the Standard Oil Company, which John D. Rockefeller created in 1879, and the United States Steel Corporation, which J. R Morgan put together in 1901 after buying out Andrew Carnegie's industrial empire.) The Sherman Act, which Congress passed in 1890, was, at first sight, a historic piece of legislation, for it proscribed any restraint of trade by an existing monopoly and any attempt to form a new monopoly. The Clayton Act, of 1914, which Congress enacted in order to outlaw a number of specific predatory tactics employed by American firms, such as price-discrimination and exclusivedealing contracts, also seemed at the time to be a potent measure.
65
66
In practice, though, the antitrust laws have proved to be a lot more accommodating than their language implies. "The successfid competitor, having been urged to compete, must not be turned upon when he wins," Judge Learned Hand wrote fifty years ago, and most courts have taken his advice, refusing to shackle, let alone dismantle, powerful corporations unless they were guilty of blatantly anticompetitive behavior. Faced with potentially hostile judges, the executive branch has traditionally deliberated very carefiffly before launching a criminal antitrust action. "The Justice Department, as an institution, is extremely conservative," Philip Verveer, a former government lawyer who, during the nineteen-eighties, helped bring the case against A.T&T, told me. "It has all manner of safeguards." In 1994, Anne Bingaman, who headed the Justice Department's antitrust division during the first Clinton Administration, reviewed all the evidence against Microsoft and decided that she could@t prove a broad monopolization case against it under the Sherman Act, which is what many of her staff were recommending. Instead, she settled for negotiating the consent decree that is at the center of the current legal dispute. From the moment Bingaman made that decision, it has faced heavy criticism. In early 1995, Judge Stanley Sporkin, of the United States District Court in Washington, D.C., rejected the settlement between the government and Microsoft, because it did@t address many of the charges against the firm, including its alleged use of vaporware. In a memorable ruling, Sporkin declared, "It is clear to this Court that if it signs the decree presented to it, the message will be that Microsoft is so powerful that neither the market nor the Government is capable of dealing with all of its monopoestic practices."
An appeals court subsequently overturned Sporkin's decision, on the ground that he didn't have the authority to block the settlement, but many critics still beheve that the government erred in agree' to such a limited consent decree. "Gates is like smallpox," Frederick Warren Boulton, a former chief economist in the Department of Justice's antitrust division, told me. "You have to go in there and you have to nail it. If you leave it lying around, it will just come back." Warren-Boulton, who worked for Novell, a Microsoft rival, in trying to persuade the government to sue Microsoft, compared Bingaman's decision with the Bush Administration's failure to carry on the Gulf War. "They were within two days of the capital, they could have taken Baghdad, but for some reason they decided to sit there and negotiate a consent decree. Boy, was that a mistake!" he said. lintH the Justice Department's recent action, the impact of the consent decree on Microsoft's business was minimal, just as Gates predicted. Even some senior figures at the Justice Department appear to believe that the settlement that Bingaman and her colleagues reached with Microsoft wasn't ideal. "Anytime you have a retrospective dispute, one can think of things one mleht have chaneed at the time, but I do@t think it would be appropriate of me to second-guess," one official told me. "They did what they thought was right."
Like Bingaman four years ago, Joel Klein is now under pressure to bring a monopolization case against Microsoft under the Sherman Act, this time relating to the company's behavior in the market for Internet browsers. Last month, the Justice Department hired David Boles, a top antitrust litigator, as a consultant, but Klein insisted that he has not yet made a decision on whether to go ahead with the case. "It's too early to tell," he said. Despite his ongoing tussle with Gates, Klein is keen to avoid being portrayed as a "trust-buster" in the tradition of WilEam Howard Taft and Theodore Roosevelt. "There are some people out there who have a political view that the aggregation of economic power, in and of itself, is a bad thing. I do@t think that is the role of the antitrust department," he told me. "Ours is a much more careful, market-driven approach, one that takes into account the potential competitive benefits as well as the potential competitive harms of bigness." Klein stressed that he does@t want the Justice Department to be seen "as the regulatory bureau for the Microsoft company," and he added that the government's aim is limited and straightforward: "We are trying to have a fair fight for these new
66
67 markets." I asked Klein how it felt to be taking on the most popular businessman in the countrr-a man who has become an icon for the global triumph of capitalism at the end of the twentieth century.
"Microsoft has been a tremendous American success story," he replied. "I say that publicly, and I dor;t think there is any doubt about it. This is not about good guys and bad guys."
One of the ironies of Judge Jackson's preliminary ruling is that the much-maligned consent decree could turn out to have more bite than its critics anticipated. In delaying his ultimate decision for several months, the judge has provided time for a long-overdue discussion about what the government's role should be in an information-based economy. Until now, such a debate had been prevented by the n@ive libertarian sentiments of many people in the computer world, and by a more legitimate worry that too much government intrusion could threaten the high-tech sector's remarkable record of innovation. Predictably, perhaps, economists are divided in their views on how far the government satid go. Some of them, including Steven Salop, think the forces leading tow,ard monopolization are so strong in the computer industry that fidl-scale federal regulation may eventuaUy be unavoidable. Others, including Garth Saloner and Richard Schmalensee, of M.I.T, who has acted as a consultant for Microsoft, believe that vigilant enforcement of the existing antitrust laws is the best solution available.
"The notion that you ought to replace faith in the market with faith in the government-I don't think the new theories let you do that," Schmalensee said. "They offer less reason to have unqualified faith in the market, but they don't make the government's job easier." Even Brian Arthur, whose pioneering work helped spark the revival of antitrust economics, concedes that he is of two minds about the Microsoft battle. "The wrong type of regulation could turn the high-tech sector into something like the high-tech sector in Europe or Japan, not the wild and wonderfull free-for-all that it is now," he told me. "I think America has an absolutely wonderful record of innovation in high technology, and I would hate to see that hampered." Despite these fears, Arthur believes that the Justice Department did the right thing in challenging Microsoft and trying to establish the simple but important principle that companies should have equal access to new markets. "Exactly how you do that in a positive way I don't know," he said. "But it would certainly rule out what is going on now, with Microsoft saying, 'I command a user base of millions, and I'm just going to lever them over into the next market that I am taking over.' It's impossible to start everybody absolutely equal, but when things become egregious the government ought to step in.'
67
68
Appeared in Harvard Business Review, July-Aug.,1996
Increasing Returns and
the New World of Business
by W. Brian Arthur *
April 27, 1996
68
69
Our understanding of how markets and businesses operate was passed down to us more than a century ago by a handful of European economists—Alfred Marshall in England and a few of his contemporaries on the continent. It is an understanding based squarely upon the assumption of diminishing returns: products or companies that get ahead in a market eventually run into limitations, so that a predictable equilibrium of prices and market shares is reached. The theory was in rough measure valid for the bulk-processing, smokestack economy of Marshall’s day. And it still thrives in today’s economics textbooks. But steadily and continuously in this century, Western economies have undergone a transformation from bulk-material manufacturing to design and use of technology—from processing of resources to processing of information, from application of raw energy to application of ideas. As this shift has taken place, the underlying mechanisms that determine economic behavior have shifted from ones of diminishing to ones of increasing returns. Increasing returns are the tendency for that which is ahead to get farther ahead, for that which loses advantage to lose further advantage. They are mechanisms of positive feedback that operate—within markets, businesses, and industries—to reinforce that which gains success or aggravate that which suffers loss. Increasing returns generate not equilibrium but instability: If a product or a company or a technology—one of many competing in a market—gets ahead by chance or clever strategy, increasing returns can magnify this advantage, and the product or company or technology can go on to lock in the market. More than causing products to become standards, increasing returns cause businesses to work differently, and they stand many of our notions of how business operates on their head. Mechanisms of increasing returns exist alongside those of diminishing returns in all industries. But roughly speaking, diminishing returns hold sway in the traditional part of the economy—the processing industries. Increasing returns reign in the newer part—the knowledge-based industries. Modern economies have therefore become divided into two interrelated, intertwined parts—two worlds of business—corresponding to the two types of returns. The two worlds have different economics. They differ in behavior, style, and culture. They call for different management techniques, different strategies, different codes of government regulation. They call for different understandings.
Alfred Marshall’s World
Let’s go back to beginnings—to the diminishing-returns view of Alfred Marshall and his contemporaries. Marshall’s world of the 1880s and 1890s was one of bulk production: of metal ores, aniline dyes, pig iron, coal, lumber, heavy chemicals, soybeans, coffee—commodities heavy on resources, light on know-how. In that world it was reasonable to suppose, for example, that if a coffee plantation expanded production it would ultimately be driven to use land less suitable for coffee—it would run into diminishing returns. So if coffee plantations competed, each would expand until it ran into limitations in the form of rising costs or diminishing profits. The market would be shared by many plantations, and a market price would be established at a predictable level—depending on tastes for coffee and the availability of suitable farmland. Planters would produce coffee so long as doing so was profitable, but because the price would be squeezed down to the average cost of production, no one would make a killing. Marshall said such a market was in perfect competition, and the economic world he envisaged fitted beautifully with the Victorian values of his time. It was at equilibrium and therefore orderly, predictable and therefore amenable to scientific analysis, stable and therefore safe, slow to change and therefore continuous. Not too rushed, not too profitable. In a word, mannerly. In a word, genteel. With a few changes, Marshall’s world lives on a century later within that part of the modern economy still devoted to bulk processing: of grains, livestock, heavy chemicals, metals and ores, foodstuffs, retail goods—the part where operations are largely repetitive day to day or week to week. Product differentiation and brand names now mean that a few companies rather than many compete in a given market. But typically, if these companies try to expand, they run into some limitation: in numbers of consumers who prefer their brand, in regional demand, in access to raw materials. So no company can corner the market. And because such products are normally substitutable for one another, something
69 70 like a standard price emerges. Margins are thin and nobody makes a killing. This isn’t exactly Marshall’s perfect competition, but it approximates it.
The Increasing-Returns World
What would happen if Marshall’s diminishing returns were reversed so that there were increasing returns? If products that got ahead thereby got further ahead, how would markets work? Let’s look at the market for operating systems for personal computers in the early 1980s when CP/M, DOS, and Apple’s Macintosh systems were competing. Operating systems show increasing returns: If one system gets ahead, it attracts further software developers and hardware manufacturers to adopt it, which helps it get further ahead. CP/M was first in the market and by 1979 was well established. The Mac arrived later but was wonderfully easy to use. DOS was born when Microsoft locked up a deal in 1980 to supply an operating system for the IBM PC. For a year or two, it was by no means clear which system would win. The new IBM PC—DOS’s platform—was a kludge. But the growing base of DOS/IBM users encouraged software developers such as Lotus to write for DOS. DOS’s prevalence—and the IBM PC’s—bred further prevalence, and eventually the DOS/IBM combination came to dominate a large portion of the market. That history is well known. But notice several things: It was not predictable in advance (before the IBM deal) which system would come to dominate. Once DOS/IBM got ahead it locked in the market because it did not pay users to switch. The dominant system was not the best: DOS was derided by computer professionals. And once DOS locked in the market, its sponsor Microsoft was able to spread its costs over a large base of users—it enjoyed killer margins. These properties, then, have become the hallmarks of increasing returns: market instability (the market tilts to favor a product that gets ahead), multiple potential outcomes (under different events in history different operating systems could have won), unpredictability, the ability to lock in a market, the possible predominance of an inferior product, and fat profits for the winner. They surprised me when I first perceived them in the late 1970s. They were also repulsive to economists brought up on the order, predictability, and optimality of Marshall’s world. Glimpsing some of these properties in 1939, English economist John Hicks warned that admitting increasing returns would lead to “the wreckage of the greater part of economic theory.” But Hicks had it wrong: the theory of increasing returns does not destroy the standard theory—it complements it. Hicks felt repugnance not just because of unsavory properties, but because in his day no mathematical apparatus existed to analyze increasing-returns markets. That situation has now changed. Using sophisticated techniques from qualitative dynamics and probability theory, I and others have developed methods to analyze increasing returns markets. The theory of increasing returns is new, but it already is well established. And it renders such markets amenable to economic understanding. In the early days of my work on increasing returns, I was told they were an anomaly. Like some exotic particle in physics, they might exist in theory but would be rare in practice. And if they did exist, they would last for only a few seconds before being arbitraged away. But by the mid-1980s, I realized increasing returns were neither rare nor ephemeral. In fact, a major part of the economy in fact was subject to increasing returns—high technology.
Why should this be so? Several reasons: Up-front Costs. High-tech products—pharmaceuticals, computer hardware and software, aircraft and missiles, telecommunications equipment, bioengineered drugs, and suchlike—are by definition complicated to design and to deliver to the market place. They are heavy on know-how and light on resources. Hence they typically have R&D costs that are large relative to their unit production costs. The first disk of Windows to go out the door cost Microsoft $50M, the second and subsequent disks cost $3. Unit costs fall as sales increase.
Network Effects. Many high-tech products need to be compatible with a network of users. So if much downloadable software on the Internet will soon appear as programs written in Sun Microsystems’ Java language, users will need Java on their computers to run them. Java has competitors. But the more it gains prevalence, the more likely it will emerge as a standard.
70 71
Customer Groove-In. High tech products are typically difficult to use. They require training. Once users invest in this training—say the maintenance and piloting of Airbus passenger aircraft—they merely need to update these skills for subsequent versions of the product. As more market is captured, it becomes easier to capture future markets. In high-tech markets, such mechanisms ensure that products that gain market advantage stand to gain further advantage, making these markets unstable and subject to lock-in. Of course, lock-in is not forever. Technology comes in waves, and a lock-in, such as DOS’s, can only last as long as a particular wave lasts. So, we can usefully think of two economic regimes or worlds: a bulk-production world yielding products that essentially are congealed resources with a little knowledge and operating according to Marshall’s principles of diminishing returns, and a knowledge-based part of the economy yielding products that essentially are congealed knowledge with a little resources and operating under increasing returns. The two worlds are not neatly split. Hewlett- Packard, for example, designs knowledge-based devices in Palo Alto, California, and manufactures them in bulk in places like Corvallis, Oregon or Greeley, Colorado. Most high-tech companies have both knowledge-based operations and bulk-processing operations. But because the rules of the game are different for each, companies often separate them—as Hewlett-Packard does. Conversely, manufacturing companies have operations such as logistics, branding, marketing, and distribution that belong largely to the knowledge world. And some products—like the IBM PC—start in the increasing returns world, but later in their life cycle become virtual commodities that belong to Marshall’s processing world.
The Halls of Production and the Casino of Technology
Because the two worlds of business—processing bulk goods, and crafting knowledge into products—differ in their underlying economics, it follows that they differ in their character of competition and their culture of management. It is a mistake to think that what works in one world is appropriate for the other. There is much talk these days about a new management style that involves flat hierarchies, mission orientation, flexibility in strategy, market positioning, reinvention, restructuring, reengineering, repositioning, reorganization, and re-everything else. Are these new insights, or are they fads? Are they appropriate for all organizations? Why are we seeing this new management style? Let us look at the two cultures of competition. In bulk processing, a set of standard prices typically emerges. Production tends to be repetitive—much the same from day to day or even from year to year. Competing therefore means keeping product flowing, trying to improve quality, getting costs down. There is an art to this sort of management, one widely discussed in the literature. It favors an environment free of surprises or glitches—an environment characterized by control and planning. Such an environment requires not just people to carry out production but people to plan and control it. So it favors a hierarchy of bosses and workers. Because bulk processing is repetitive, it allows constant improvement, constant optimization. And so, Marshall’s world tends to be one that favors hierarchy, planning, controls. Above all, it is a world of optimization. Competition is different in knowledge-based industries, because the economics are different. If knowledge-based companies are competing in winner-take-most markets, then managing becomes redefined as a series of quests for the next technological winner—the next cash cow. The goal becomes the search for the Next Big Thing. In this milieu, management becomes not production oriented but mission oriented. Hierarchies flatten not because democracy is suddenly bestowed on the work force or because computers can cut out much of middle management. They flatten because, to be effective, the deliverers of the next-thing-for-the-company need to be organized like commando units in small teams that report directly to the CEO or to the board. Such people need free rein. The company’s future survival depends upon them. So they—and the commando teams that report to them in turn—will be treated not as employees but as equals in the business of the company’s success. Hierarchy dissipates and dissolves.
71 72
Does this mean hierarchy should disappear in meatpacking, steel production, or the navy? Contrary to recent management evangelizing, a style that is called for in Silicon Valley will not necessarily work in the processing world. An aircraft’s safe arrival depends on the captain, not the flight attendants. The cabin crew can usefully be “empowered” and treated as human beings. This is wise and proper. But forever there will be a distinction—a hierarchy—between cockpit and cabin crews. In fact, the style in the diminishing-returns Halls of Production is much like that of a sophisticated modern factory: the goal is to keep high-quality product flowing at low cost. There is little need to watch the market every day, and when things are going smoothly, the tempo can be leisurely. By contrast, the style of competition in the increasing returns arena is more like gambling. Not poker, where the game is static and the players vie for a succession of pots. It is casino gambling, where part of the game is to choose which games to play, as well as playing them with skill. We can imagine the top figures in high tech—the Gateses and Gerstners and Groves of their industries—as milling in a large casino. Over at this table, a game is starting called multimedia. Over at that one, a game called Web services. In the corner is electronic banking. There are many such tables. You sit at one. How much to play? you ask. Three billion, the croupier replies. Who’ll be playing? We won’t know until they show up. What are the rules? Those’ll emerge as the game unfolds. What are my odds of winning? We can’t say.
Do you still want to play? High tech, pursued at this level, is not for the timid. In fact, the art of playing the tables in the Casino of Technology is primarily a psychological one. What counts to some degree—but only to some degree—is technical expertise, deep pockets, will, and courage. Above all, the rewards go to the players who are first to make sense of the new games looming out of the technological fog, to see their shape, to cognize them. Bill Gates is not so much a wizard of technology as a wizard of precognition, of discerning the shape of the next game. We can now begin to see that the new style of management is not a fad. The knowledge-based part of the economy demands flat hierarchies, mission orientation, above all a sense of direction. Not five-year plans. We can also fathom the mystery of what I’ve alluded to as re-everything. Much of this “re-everything” predilection—in the bulk-processing world—is a fancy label for streamlining, computerizing, downsizing. However, in the increasing-returns world, especially in high tech, re-everything has become necessary because every time the quest changes the company needs to change. It needs to reinvent its purpose, its goals, its way of doing things. In short, it needs to adapt. And adaptation never stops. In fact, in the increasing-returns environment I’ve just sketched, standard optimization makes little sense. You cannot optimize in the casino of increasing-returns games. You can be smart. You can be cunning. You can position. You can observe. But when the games themselves are not even fully defined, you cannot optimize. What you can do is adapt. Adaptation, in the proactive sense, means watching for the next wave that is coming, figuring out what shape it will take, and positioning the company to take advantage of it. Adaptation is what drives increasing-returns businesses, not optimization.
Playing the High-Tech Tables
Suppose you are a player in the knowledge-industry casino, in this increasing-returns world. What can you do to capitalize on the increasing returns at your disposal? How can you use them to capture markets? What strategic issues do you need to think about? In the processing world, strategy typically hinges upon capitalizing on core competencies, pricing competitively, getting costs down, bringing quality up. These are important also in the knowledge-based world, but so too are other strategies that make use of the special economics of positive feedbacks.
Two maxims are widely accepted in knowledge-based markets: it pays to hit the market first, and it pays to have superb technology. These maxims are true, but they do not guarantee success. Prodigy was first into the on-line services market, but was passive in building its subscriber base to take advantage of increasing returns. As a result, it has fallen from its leading position and currently lags the other services. As for technology, Steve Jobs’s NeXT workstation was
72 73 superb. But it was launched into a market already dominated by Sun Microsystems and Hewlett-Packard. It failed. A new product often needs to be twice or three times better in some dimension—price, speed, convenience—to dislodge a locked-in rival. So in knowledge-based markets, entering first with a fine product can yield advantage. But as strategy this is still too passive. What is needed is active management of increasing returns. One active strategy is to discount heavily initially to build up installed base. Netscape handed out its Internet browser for free and won 70% of its market. Now it can profit from spin-off software and applications. Although such discounting is effective—and widely understood—it is not always implemented. Companies often err by pricing high initially to recoup expensive R&D costs. Yet even smart discounting to seed the market is ineffective unless the resulting installed base is exploited later. America Online built up a lead of more than 4.5 million subscribers by giving away free services. But because of the Internet’s dominance, it is not yet clear it can transform this huge base into later profits. Let’s get a bit more sophisticated. Technological products do not stand alone. They depend on the existence of other products and other technologies. The Internet’s World Wide Web operates within a grouping of businesses that include browsers, on-line news, E-mail, network retailing, and financial services. Pharmaceuticals exist within a network of physicians, testing labs, hospitals, and HMO’s. Laser printers are part of a grouping of products that include computers, publishing software, scanners, and photo-input devices. Unlike products of the processing world, such as soybeans or rolled steel, technological products exist within local groupings of products that support and enhance them. They exist in mini-ecologies.
This interdependence has deep implications for strategy. When in the mid-1980s Novell introduced its network operating system, Netware, as a way of connecting personal computers in local networks, Novell made sure that Netware was technically superior to its rivals. It also heavily discounted Netware to build installed base. But these tactics were not enough. Novell recognized that Netware’s success depended on attracting software applications to run on Netware—which was a part of the ecology outside the company’s control. So it set up incentives for software developers to write for Netware rather than for its rivals. The software writers did just that. And by building Netware’s success they ensured their own. Novell managed these cross-product positive feedbacks actively to lock in its market. It went on to profit hugely from upgrades, spin-offs, and applications of its own. Another strategy that uses ecologies is linking and leveraging. This means transferring a user base built up upon one node of the ecology (one product) to neighboring nodes or products. The strategy is much like that in the game Go: you surround neighboring markets one by one, lever your user base onto them, and take them over—all the time enhancing your position in the industry. Microsoft levered its 60 million person user base in DOS onto Windows, then onto Windows 95, and then onto Microsoft Network by offering cheap upgrades and by bundling applications. The strategy has been challenged legally. But it recognizes that positive feedbacks apply across markets as well as within markets.
In fact, if technological ecologies are now the basic units for strategy in the knowledge-based world, players compete not by locking in a product on their own but by building webs—loose alliances of companies organized around a mini-ecology—that amplify positive feedbacks to the base technology. Apple, in closing its Macintosh system to outsiders in the 1980s, opted not to create such a web. It believed that with its superior technology, it could hold its increasing-returns market to itself. Apple indeed dominates its Mac-based ecology. But this ecology is now only 8% of the personal computer business. IBM erred in the other direction. By passively allowing other companies to join its PC web as clones, IBM achieved a huge user base and locked in the market. But IBM itself wound up with a small share of the spoils. The key in web building is active management of the cross-company mutual feedbacks. This means making a careful choice of partners to build upon. It also means that rather than attempting to take over all products in the ecology, dominant players in a web should allow dependent players to lock in their dependent products by piggybacking on the web’s success. By thus ceding some of the profits, dominant players ensure that all participants remain committed to the alliance. Important also to strategy in knowledge-based markets is psychological positioning. Under increasing returns, rivals will back off in a market not only if it is locked in but if they believe it will be locked in by someone else. Hence
73 74 we see psychological jockeying in the form of preannouncements, feints, threatened alliances, technological preening, touted future partnerships, parades of vaporware (announced products that don’t yet exist). This posturing and puffing acts much as similar behavior does in a primate colony: it discourages competitors from taking on a potentially dominant rival. No moves need be made in this strategy of pre-market facedown. It is purely a matter of psychology. What if you hold a losing hand? Sometimes it pays to hold on for residual revenue. Sometimes a fix can be provided by up-dated technology, fresh alliances, product changes. But usually under heavy lock-in, these tactics do not work. The alternatives are then slow death or graceful exit—relinquishing the field to concentrate on positioning for the next technology wave. Exit may not mean quitting the business entirely. America Online, Compuserve, Prodigy, and Microsoft Network have all ceded dominance of the on-line computer networking market to the Internet. But instead of exiting, they are steadily becoming adjuncts of the Net, supplying content services such as financial quotations or games and entertainment. They have lost the main game. But they will likely continue in a side game with its own competition for dominance within the Net’s ecology. Above all, strategy in the knowledge world requires CEOs to recognize that a different kind of economics is at work. CEOs need to understand which positive and negative feedback mechanisms are at play in the market ecologies they compete in. Often there are several such mechanisms—interbraided, operating over different time frames, each needing to be understood, observed, and actively managed.
What about Service Industries?
So far, I’ve talked mainly about high tech. Where do service industries such as insurance, restaurants, and banking fit in? Which world do they belong to? The question is tricky. It would appear that such industries belong to the diminishing-returns, processing part of the economy because often there are regional limits to the demand for a given service, most services do consist of “processing” clients, and services are low-tech. The truth is that network or user-base effects often operate in services. Certainly, retail franchises exist because of increasing returns. The more McDonalds’s restaurants or Motel 6 franchises are out there geographically, the better they are known. Such businesses are patronized not just for their quality but because people want to know exactly what to expect. So the more prevalent they are, the more prevalent they can become. Similarly, the larger a bank’s or insurance company’s customer base, the more it can spread its fixed costs of headquarters staff, real estate, and computer operations. These industries, too, are subject to mild increasing returns. So we can say more accurately that service industries are a hybrid. From day to day, they act like bulk-processing industries. But over the long-term, increasing returns will dominate—even though their destabilizing effects are not as pronounced as in high tech. The U.S. airline business, for example, processes passengers day to day. So it seemed in 1981 that deregulation should enhance competition, as it normally does under diminishing returns. But over the long term, airlines in fact experience a positive feedback: under the hub-and-spoke system, once an airline gets into trouble it cannot work the feeder system for its routes properly, its fleet ages, it starts a downward spiral, and it loses further routes. The result of deregulation over the long term has been a steady decline in large carriers, from 15 in 1981 to around 6 at present. Some routes have become virtual monopolies, with resulting higher fares. None of this was intended. But it should have been predicted—given increasing returns. In fact, the increasing-returns character of service industries is steadily strengthening. One of the marks of our time is that in services everything is going software—everything that is information based. So operations that were once handled by people—designing fancy financial instruments or cars or fashion goods, processing insurance claims, supplying and inventorying in retail, conducting paralegal searches for case precedents—are increasingly being handled by software. As this reengineering of services plays out, centralized software facilities come to the fore. Service providers become hitched into software networks, regional limitations weaken, and user-base network effects kick in. This phenomenon can have two consequences. First, where the local character of service remains important, it can preserve a large number of service companies, but clustered round a dominant software provider—like the large
74 75 numbers of small, independent law firms tied in to the dominant computer-search network, Lexis-Nexis. Or physicians tied in to an HMO. Second, where locality is unimportant, network effects can transform competition toward the winner-take-most character we see in high tech. For example, when Internet-based retail banking arrives, regional demand limitations will vanish. Each virtual bank will gain in advantage as its network increases. Barring regulation, consumer banking will then become a contest among a few large banking networks. It will become an increasing returns business.
Services belong to both the processing and the increasing-returns world. But their center of gravity is crossing over to the latter.
In the Case of Microsoft…
What should be legal in this powerful and as yet unregulated world of increasing returns? What constitutes fair play? Should technology markets be regulated, and if so in what way? These questions have come to a head with the huge publicity generated by the US Justice Department’s current antitrust case against Microsoft. In Marshall’s world, antitrust regulation is well understood. Allowing a single player to control, say, more than 35% of the silver market is tantamount to allowing monopoly pricing, and the government rightly steps in. In the increasing returns world, things are more complicated. There are arguments in favor of allowing a product or company in the web of technology to dominate a market, as well as arguments against. Consider these pros and cons:
Convenience. A locked-in product may provide a single standard of convenience. If a software company such as Microsoft allows us to double-click all the way from our computer screen straight to our bank account (by controlling all the technologies in between), this avoids a tedious balkanizing of standards, where we have to spend useless time getting into a succession of on-line connection products. Fairness. If a product locks-in a market because it is superior, this is fair, and it would be foolish to penalize such success. If it locks-in merely because user-base was levered over from a neighboring lock-in, this is unfair.
Technology development: A locked-in product may obstruct technological advancement. If a clunker such as DOS locks up the PC market for 10 years, there is little incentive for other companies to develop alternatives. The result is impeded technological progress.
Pricing: To lock in, a product usually has been discounted, and this established low price is often hard to raise later. So monopoly pricing—of great concern in bulk processing markets—is therefore rarely a major worry. Added to these considerations, high tech is not a commodity industry. Dominance may not so much consist in cornering a single product as in successively taking over more and more threads of the web of technology, thereby preventing other players from getting access to new, breaking markets. It would be difficult to separate out each thread and to regulate it. And of course it may be impracticable to regulate a market before it forms—before it is even fully defined. There are no simple answers to antitrust regulation in the increasing returns world. On balance, I would favor a high degree of regulatory restraint, with two key principles: Do not penalize success. Short term monopolization of an increasing returns market is correctly perceived as a reward or prize for innovation and risk taking. There is a temptation to single out dominant players and hit them with an antitrust suit. This reduces regulation to something like a brawl in a old-West saloon—if you see a head, hit it. Not a policy that preserves an incentive to innovate in the first place.
No head starts for the privileged. This means that as a new market opens up, such as electronic consumer banking, companies that already dominate standards, operating systems, and neighboring technologies should not be allowed a ten-mile start in the land-rush that follows. All competitors should have fair and open access to the applicable technologies and standards. In practice, these principles would mean allowing the possibility of winner-take-all jackpots in each new sub- industry, in each new wave of technology. But each contender should have access to whatever degree possible to the
75 76 same technologies, the same open standards, so that all are lined up behind the same starting line. If industry does not make such provisions voluntarily, government regulation will.
Thoughts for Managers
Where does all this leave us? At the beginning of this century, industrial economies were based largely on the bulk processing of resources. At the close of the century, they are based on the processing of resources and on the processing of knowledge. Economies have bifurcated into two worlds—intertwined, overlapping, and different. These two worlds operate under different economic principles. Marshall’s world is characterized by planning, control, and hierarchy. It is a world of materials, of processing, of optimization. The increasing returns world is characterized by observation, positioning, flattened organizations, missions, teams, and cunning. It is a world of psychology, of cognition, of adaptation.
Many managers have some intuitive grasp of this new increasing returns world. Few understand it thoroughly. Here are some questions managers need to ask themselves when they operate in knowledge-based markets: Do I understand the feedbacks in my market? In the processing world, understanding markets means understanding consumers’ needs, distribution channels, and rivals’ products. In the knowledge world, success requires a thorough understanding of the self-negating and self-reinforcing feedbacks in the market—the diminishing and increasing returns mechanisms. These feedbacks are interwoven and operate at different levels in the market and over different time frames. Which ecologies am I in? Technologies exist not alone but in an interlinked web, or ecology. It is important to understand the ecologies a company’s products belong to. Success or failure is often decided not just by the company but by the success or failure of the web it belongs to. Active management of such a web can be an important magnifier of increasing returns. Do I have the resources to play? Playing one of the increasing returns games in the casino of technology requires several things: excellent technology, the ability to hit the market at the right time, deep pockets, strategic pricing, and a willingness to sacrifice current profits for future advantage. All this is a matter not just of resources, but also of courage, resolution, will. And part of that resolution, that courage, is also the decisiveness to leave the market when increasing returns are moving against one. Hanging onto a losing position that is being further eroded by positive feedbacks requires throwing reinforcements into a battle already lost. Better to exit with financial dignity. What games are coming next? Technology comes in successive waves. Those who have lost out on this wave can position for the next. Conversely, those who have made a killing on this cycle should not become complacent. The ability to profit under increasing returns is only as good as the ability to see what’s coming in the next cycle, and to position oneself for it—technologically, psychologically, and cooperatively. In high tech, it is as if we are moving slowly on a ship, with new technologies looming, taking shape, through a fog of unknowingness. Success goes to those who have the vision to foresee, to imagine, what shapes these next games will take. These considerations appear daunting. But increasing returns games provide large payoffs for those brave enough to play them and win. And they are exciting. Processing, in the service or manufacturing industries, has its own risks. Precisely because processing is low margin, operations must struggle to stay afloat. Neither world of business is for the fainthearted.
Technology thinker George Gilder has remarked, “The central event of the twentieth century is the overthrow of matter. In technology, economics, and the politics of nations, wealth in the form of physical resources is steadily declining in value and significance. The powers of mind are everywhere ascendant over the brute force of things.” As the economy shifts steadily away from the brute force of things into the powers of mind, from resource-based bulk processing into knowledge-based design-and-reproduction, so it is shifting from a base of diminishing-returns to one of increasing-returns. A new economics—one very different from that in the textbooks—now applies, and nowhere is this more true than in high technology. Success will strongly favor those who understand this new way of thinking.
76 77
W. Brian Arthur is Citibank Professor at the Santa Fe Institute and Dean and Virginia Morrison Professor of Economics and Population Studies, at Stanford University. He is the author of Increasing Returns and Path Dependence in the Economy, University of Michigan Press, Ann Arbor, Michigan, 1994.
His webpage is: http://santafe.edu/arthur Acknowledgements The author is indebted to Eric Beinhocker, Larry Blume, John Casti, David Lane, Cormac McCarthy, Martin Shubik, and Lynda Woodman for discussions of the ideas in this paper.
77 78
From The Economy as an Evolving Complex System II
Edited by W. Brian Arthur, Steven Durlauf and David Lane, Addison-Wesley, Reading, Mass., 1997
Introduction: Process and Emergence in the Economy by W. Brian Arthur, Steven Durlauf and David A. Lane. In September 1987 twenty people came together at the Santa Fe Institute to talk about "the economy as a evolving, complex system." Ten were theoretical economists, invited by Kenneth J. Arrow, and ten were physicists, biologists and computer scientists, invited by Philip W. Anderson. The meeting was motivated by the hope that new ideas bubbling in the natural sciences, loosely tied together under the rubric of "the sciences of complexity," might stimulate new ways of thinking about economic problems. For ten days, economists and natural scientists took turns talking about their respective worlds and methodologies. While physicists grappled with general equilibrium analysis and noncooperative game theory, economists tried to make sense of spin glass models, Boolean networks, and genetic algorithms. The meeting left two legacies. The first was a volume of essays, The Economy as an Evolving Complex System, edited by Arrow, Anderson and David Pines. The other was the founding, in 1988, of the Economics Program at the Santa Fe Institute, the Institute's first resident research program. The Program's mission was to encourage the understanding of economic phenomena from a complexity perspective, which involved the development of theory as well as tools for modeling and for empirical analysis. To this end, since 1988, the Program has brought researchers to Santa Fe, sponsored research projects, held several workshops each year, and published several dozen working papers. And since 1994, it has held an annual summer school for economics graduate students. This volume, The Economy as an Evolving Complex System II, represents the proceedings of an August, 1996 workshop sponsored by the SFI Economics Program. The intention of this workshop was to take stock, to ask: What has a complexity perspective contributed to economics in the past decade? In contrast to the 1987 workshop, almost all of the presentations addressed economic problems, and most presenters were economists by training. In addition, while some of the work presented was conceived or carried out at the Institute, some of the participants had no previous relation with SFI--research related to the complexity perspective is under active development now in a number of different institutes and university departments. But just what is the complexity perspective in economics? That is not an easy question to answer. Its meaning is still very much under construction, and in fact the present volume is intended to contribute to that construction process. Indeed, the authors of the essays in this volume by no means share a single, coherent vision of the meaning and significance of complexity in economics. What we will find instead is a family resemblance, based upon an interrelated set of themes that together constitute the current meaning of the complexity perspective in economics. Several of these themes, already active subjects of research by economists in the mid-1980s, are well described in the earlier Economy as an Evolving Complex System : in particular, applications of nonlinear dynamics to economic theory and data analysis, surveyed in the 1987 meeting by Michele Boldrin and William Brock; and the theory of positive feedback and its associated phenomenology of path-dependence and lock-in, discussed by W. Brian Arthur. Research related to both these themes has flourished since 1987, both in and outside the SFI Economics Program. While chaos has been displaced from its 1987 place at center stage of the interest in nonlinear dynamics, in the last decade economists have made substantial progress in
78 79 identifying patterns of nonlinearity in financial time series and in proposing models that both offer explanations for these patterns and help to analyze and even to some extent predict the series in which they are displayed. Brock surveys both these developments in his chapter in this volume, while positive feedback plays a central role in the models analyzed by Lane (on information contagion), Durlauf (on inequality) and Krugman (on economic geography), and lurk not far under the surface of the phenomena described by North (development) and Leijonhufvud (high inflation). Looking back over the developments in the past decade, and of the papers produced by the program, we believe that a coherent perspective--sometimes called the "Santa Fe approach"--has emerged within economics. We will call this the complexity perspective, or Santa Fe perspective, or occasionally the process-and-emergence perspective. Before we describe this, we first sketch the two conceptions of the economy that underlie standard, neoclassical economics (and indeed most of the presentations by economic theorists at the earlier, 1987 meeting). We can call these conceptions the "equilibrium" and "dynamical systems" approaches. In the equilibrium approach, the problem of interest is to derive, from the rational choices of individual optimizers, aggregate-level "states of the economy" (prices in general equilibrium analysis, a set of strategy assignments in game theory with associated payoffs) that satisfy some aggregate-level consistency condition (market-clearing, Nash equilibrium), and to examine the properties of these aggregate-level states. In the dynamical systems approach, the state of the economy is represented by a set of variables, and a system of difference equations or differential equations describes how these variables change over time. The problem is to examine the resulting trajectories, mapped over the state space. However, the equilibrium approach does not describe the mechanism whereby the state of the economy changes over time nor indeed how an equilibrium comes into being. [1] And the dynamical system approach generally fails to accommodate the distinction between agent- and aggregate-levels (except by obscuring it through the device of "representative agents"). Neither accounts for the emergence of new kinds of relevant state variables, much less new entities, new patterns, new structures. [2] To describe the complexity approach, we begin by pointing out six features of an economy that together present difficulties for the traditional mathematics used in economics: [3] Dispersed Interaction What happens in the economy is determined by the interaction of many dispersed, possibly heterogeneous, agents acting in parallel. The action of any given agent depends upon the anticipated actions of a limited number of other agents and on the aggregate state these agents co-create No Global Controller No global entity controls interactions. Instead, controls are provided by mechanisms of competition and coordination between agents. Economic actions are mediated by legal institutions, assigned roles, and shifting associations. Nor is there a universal competitor--a single agent that can exploit all opportunities in the economy Cross-cutting Hierarchical Organization The economy has many levels of organization and interaction. Units at any given level behaviors, actions, strategies, products typically serve as `building blocks' for constructing units at the next higher level. The overall organization is more than hierarchical, with many sorts of tangling interactions (associations, channels of communication) across levels Continual Adaptation Behaviors, actions, strategies, and products are revised continually as the individual agents accumulate experience--the system constantly adapts Perpetual Novelty Niches are continually created by new markets, new technologies, new behaviors, new institutions. The very act of filling a niche may provide new niches. The result is ongoing, perpetual novelty
79 80
Out-of-Equilibrium Dynamics Because new niches, new potentials, new possibilities, are continually created, the economy operates far from any optimum or global equilibrium. Improvements are always possible and indeed occur regularly. Systems with these properties have come to be called adaptive nonlinear networks. (The term is John Holland's, 1987.) There are many such in nature and society: nervous systems, immune systems, ecologies, as well as economies. An essential element of adaptive nonlinear networks is that they do not act simply in terms of stimulus and response. Instead they anticipate. In particular, economic agents form expectations--they build up models of the economy and act on the basis of predictions generated by these models. These anticipative models need neither be explicit, nor coherent, nor mutually consistent. Because of the difficulties outlined above, the mathematical tools economists customarily use, which exploit linearity, fixed points, and systems of differential equations, cannot provide a deep understanding of adaptive nonlinear networks. Instead, what is needed is new classes of combinatorial mathematics and population-level stochastic processes, in conjunction with computer modeling. These mathematical and computational techniques are in their infancy. But they emphasize the discovery of structure and the processes through which structure emerges across different levels of organization. This conception of the economy as an adaptive nonlinear network--an evolving, complex system--has profound implications for the foundations of economic theory and for the way in which theoretical problems are cast and solved. We interpret these implications as follows: Cognitive foundations Neoclassical economic theory has a unitary cognitive foundation: economic agents are rational optimizers. This means that (in the usual interpretation) agents evaluate uncertainty probabilistically, revise their evaluations in the light of new information via Bayesian updating, and choose the course of action that maximizes their expected utility. As glosses on this unitary foundation, agents are generally assumed to have common knowledge about each other and rational expectations about the world they inhabit (and of course co-create). In contrast, the Santa Fe viewpoint is pluralistic. Following modern cognitive theory, we posit no single, dominant mode of cognitive processing. Rather, we see agents as having to cognitively structure the problems they face--as having to "make sense" of their problems--as much as solve them. And they have to do this with cognitive resources that are limited. To "make sense," to learn, and to adapt, agents use variety of distributed cognitive processes. The very categories agents use to convert information about the world into action emerge from experience, and these categories or cognitive props need not fit together coherently in order to generate effective actions. Agents therefore inhabit a world that they must cognitively interpret--one that is complicated by the presence and actions of other agents and that is ever changing. It follows that agents generally do not optimize in the standard sense, not because they are constrained by finite memory or processing capability, but because the very concept of an optimal course of action often cannot be defined. It further follows that the deductive rationality of neoclassical economic agents occupies at best a marginal position in guiding effective action in the world. And it follows that any "common knowledge" agents might have about one another must be attained from concrete, specified cognitive processes operating on experiences obtained through concrete interactions. Common knowledge cannot simply be assumed into existence. Structural foundations In general equilibrium analysis, agents do not interact with one another directly, but only through impersonal markets. By contrast in game theory, all players interact with all other players, with outcomes specified by the game's payoff matrix. So interaction structures are simple and often extreme--one-with-all or all-with-all. Moreover, the internal structure of the agents themselves is abstracted away.[4] In contrast, from a complexity perspective, structure matters. First, network-based structures become important. All economic action involves
80 81 interactions among agents, so economic functionality is both constrained and carried by networks defined by recurring patterns of interaction among agents. These network structures are characterized by relatively sparse ties. Second, economic action is structured by emergent social roles and by socially-supported procedures--that is, by institutions. Third, economic entities have a recursive structure: they are themselves comprised of entities. The resulting "level" structure of entities and their associated action processes is not strictly hierarchical, in that component entities may be part of more than one higher-level entity and entities at multiple levels of organization may interact. Thus reciprocal causation operates between different levels of organization--while action processes at a given level of organization may sometimes by viewed as autonomous, they are nonetheless constrained by action patterns and entity structures at other levels. And they may even give rise to new patterns and entities at both higher and lower levels. From the Santa Fe perspective, the fundamental principle of organization is the idea of that units at one level combine to produce units at the next higher level.[5] What counts as a problem and as a solution It should be clear by now that exclusively posing economic problems as multi-agent optimization exercises makes little sense from the viewpoint we are outlining--a viewpoint that puts emphasis on process, not just outcome. In particular, it asks how new "things" arise in the world--cognitive things, like "internal models"; physical things, like "new technologies"; social things, like new kinds of economic "units." And it is clear that if we posit a world of perpetual novelty, then outcomes cannot correspond to steady-state equilibria, whether Walrasian, Nash, or dynamic-systems-theoretical. The only descriptions that can matter in such a world are about transient phenomena--about process and about emergent structures. What then can we know about the economy from a process-and-emergence viewpoint, and how can we come to know it? Studying process and emergence in the economy has spawned a growth industry in the production of what are now generally called "agent-based models." And what counts as a solution in an agent-based model is currently under negotiation. Many of the papers in this volume--including those by Arthur et al, Darley and Kauffman, Shubik, Lindgren, Kollman et al, Kirman, and Tesfatsion--address this issue, explicitly or implicitly. We can characterize these as seeking emergent structures arising in interaction processes, in which the interacting entities anticipate the future through cognitive procedures that themselves involve interactions taking place in multi-level structures.
* * * A description of an approach to economics, however, is not a research program. To build a research program around a process-and-emergence perspective, two things have to happen. First, concrete economic problems have to be identified for which the approach may provide new insights. A number of candidates are offered in this volume: artifact innovation (Lane and Maxfield), the evolution of trading networks (Ioannides, Kirman and Tesfatsion), money (Shubik), the origin and spatial distribution of cities (Krugman), asset pricing (Arthur et al., Brock), high inflation (Leijonhuvfud), persistent differences in income between different neighborhoods or countries (Durlauf). Second, cognitive and structural foundations for modeling these problems have to be constructed, and methods developed for relating theories based on these foundations to observable phenomena (Manski). Here, while substantial progress has been made since 1987, the program is far from complete. The essays in this volume describe a series of parallel explorations of the central themes of process and emergence in an interactive world--of how to study systems capable of generating perpetual novelty. These explorations do not form a coherent whole. They are sometimes complementary, sometimes even partially contradictory. But what could be more appropriate to the
81 82
Santa Fe perspective, with its emphasis on distributed processes, emergence, and self- organization? Here are our interpretations of the research directions that seem to be emerging from this process: Cognition The central cognitive issues raised in this volume are ones of interpretation. As Shubik puts it, "the interpretation of data is critical. It is not what the numbers are, but what they mean." How do agents render their world comprehensible enough so that "information" has meaning? The two papers by Arthur, Holland, LeBaron, Palmer and Tayler and by Darley and Kauffman consider this. They explore problems in which a group of agents take actions whose effects depend on what the other agents do. The agents base their actions on expectations they generate about how other agents will behave. Where do these expectations come from? Both papers reject common knowledge or common expectations as a starting point. Indeed, Arthur et al. argue that common beliefs cannot be deduced. Because agents must derive their expectations from an imagined future that is the aggregate result of other agents' expectations, there is a self-reference of expectations that leads to deductive indeterminacy. Rather, both papers suppose that each agent has access to a variety of "interpretative devices" that single out particular elements in the world as meaningful and suggest useful actions on the basis of the "information" these elements convey. Agents keep track of how useful these devices turn out to be, discarding ones that produce bad advice and tinkering to improve those that work. In this view, economic action arises from an evolving ecology of "interpretive devices" that interact with one another through the medium of the agents that use them to generate their expectations. Arthur et al. build a theory of asset pricing upon such a view. Agents--investors--act as market statisticians. They continually generate expectational models--interpretations of what moves prices in the market and test these by trading. They discard and replace models if not successful. Expectations in the market therefore become endogenous--they continually change and adapt to a market that they together create. The Arthur et al. market settles into a rich psychology, in which speculative bubbles, technical trading and persistence of volatility emerge. The homogeneous rational expectations of the standard literature become a special case--possible in theory but unlikely to emerge in practice. Brock presents a variant of this approach, allowing agents to switch between a limited number of expectational models. His model is simpler than that of Arthur et al., but he achieves analytical results, which he relates to a variety of stylized facts about financial times series, many of which have been uncovered through the application of nonlinear analysis over the past decade.. In the world of Darley and Kauffman, agents are arrayed on a lattice, and they try to predict the behavior of their lattice neighbors. They generate their predictions via an autoregressive model, and they can individually tune the number of parameters in the model and the length of the time series they use to estimate model parameters. Agents can change parameter number or history length by steps of length 1 each period, if by doing so they would have generated better predictions in the previous period. This induces a coevolutionary "interpretative dynamics," which does not settle down to a stable regime of precise, coordinated mutual expectations. In particular, when the system approaches a "stable rational-expectations state," it tends to break down into a disordered state. They use their results to argue against conventional notions of rationality, with infinite foresight horizons and unlimited deductive capability. In his paper on high inflation, Leijonhufvud poses the same problem as Darley and Kauffman: Where should we locate agent cognition, between the extremes of "infinite-horizon optimization" and "myopic adaptation"? Leijonhufvud argues that the answer to this question is context dependent. He claims that in situations of institutional break-down like high inflation, agent cognition shifts toward the "short memory/short foresight adaptive mode." The causative relation between institutional and cognitive shifts becomes reciprocal. With the shrinking of foresight
82 83 horizons, markets for long-term loans (where long-term can mean over 15 days) disappear. And as inflation accelerates, units of accounting lose meaning. Budgets cannot be drawn in meaningful ways, the executive arm of government becomes no longer fiscally accountable to parliament, and local governments become unaccountable to national governments. Mechanisms of social and economic control erode. Ministers lose control over their bureaucracies, shareholders over corporate management. The idea that "interpretative devices" such as explicit forecasting models and technical- trading rules play a central role in agent cognition fits with a more general set of ideas in cognitive science, summarized in Clark (1996). This work rejects the notion that cognition is all "in the head." Rather, interpretive aids such as autoregressive models, computers, languages or even navigational tools (as in Hutchins, 1995) and institutions provide a "scaffolding," an external structure on which much of task of interpreting the world is off-loaded. Clark (1996) argues that the distinctive hallmark of in-the-head cognition is "fast pattern completion," which bears little relation to the neoclassical economist's deductive rationality. In this volume, North takes up this theme, describing some of the ways in which institutions scaffold interpretations of what constitutes possible and appropriate action for economic agents. Lane and Maxfield consider the problem of interpretation from a different perspective. They are particularly interested in what they call attributions of functionality: interpretations about what an artifact does. They argue that new attributions of functionality arise in the context of particular kinds of agent relationships, where agents can differ in their interpretations. As a consequence, cognition has an unavoidable social dimension. What interpretations are possible depend on who interacts with whom, about what. They also argue that new functionality attributions cannot be foreseen outside the particular generative relationships in which they arise. This unforeseeability has profound consequences for what constitutes "rational" action in situations of rapid change in the structure of agent-artifact space. All the papers mentioned so far take as fundamental the importance of cognition for economic theory. But the opposite point of view can also be legitimately defended from a process- and-emergence perspective. According to this argument, overrating cognition is just another error deriving from methodological individualism, the very bedrock of standard economic theory. How individual agents decide what to do may not matter very much. What happens as a result of their actions may depend much more on the interaction structure through which they act--who interacts with whom, according to which rules. Blume makes this point in the introduction to his paper on population games, which, as he puts it, provide a class of models that shift attention "from the fine points of individual-level decision theory to dynamics of agent interaction." Padgett makes a similar claim, though for a different reason. He is interested in formulating a theory of the firm as a locus of transformative "work," and he argues that "work" may be represented by "an orchestrated sequence of actions and reactions, the sequence of which produces some collective result (intended or not)." Hence, studying the structure of coordinated action-reaction sequences may provide insight into the organization of economic activity, without bringing "cognition" into the story at all. Padgett's paper is inspired by recent work in chemistry and biology (by Eigen and Schuster and by Fontana and Buss, among others) that are considered exemplars of the complexity perspective in these fields. Structure Most human interactions, even those taking place in "economic" contexts, have a primarily social character: talking with friends, asking advice from knowledgeable acquaintances, working together with colleagues, living next to neighbors. Recurring patterns of such social interactions bind agents together into networks. [6] According to standard economic theory, what agents do depends on their values and available information. But standard theory typically ignores where values and information come from. It treats agents' values and information as exogenous and autonomous. In reality, agents learn from each other, and their values may be influenced by others'
83 84 values and actions. These processes of learning and influencing happen through the social interaction networks in which agents are embedded, and they may have important economic consequences. For example, one of the models presented in Durlauf's paper implies that value relationships among neighbors can induce persistent income inequalities between neighborhoods. Lane examines a model in which information flowing between agents in a network determines the market shares of two competing products. Kirman's paper reviews a number of models that derive economic consequences from interaction networks.
Ioannides, Kirman and Tesfatsion consider the problems of how networks emerge from initially random patterns of dyadic interaction and what kinds of structure the resulting networks exhibit. Ioannides studies mathematical models based on controlled random fields, while Tesfatsion works in the context of a particular agent-based model, in which the "agents" are strategies that play Prisoner's Dilemma with one another. Ioannides and Tesfatsion are both primarily interested in networks involving explicitly economic interactions, in particular trade. Their motivating idea, long recognized among sociologists (for example, Baker, 1984), is that markets actually function by means of networks of traders, and what happens in markets may reflect the structure of these networks, which in turn may depend on how the networks emerge.
Local interactions can give rise to large-scale spatial structures. This phenomenon is investigated by several of the papers in this volume. Lindgren's contribution is particularly interesting in this regard. Like Tesfatsion, he works with an agent-based model in which the agents code strategies for playing 2-person games. In both Lindgren's and Tesfatsion's models, agents adapt their strategies over time in response to their past success in playing against other agents. Unlike Tesfatsion's agents, who meet randomly and decide whether or not to interact, Lindgren's agents only interact with neighbors in a prespecified interaction network. Lindgren studies the emergence of spatio-temporal structure in agent space--metastable ecologies of strategies that maintain themselves for many agent- generations against "invasion" by new strategy types or "competing" ecologies at their spatial borders. In particular, he compares the structures that arise in a lattice network, in which each agent interacts with only a few other agents, and with those that arise in a fully-connected network, in which each agent interacts with all other agents. He finds that the former "give rise to a stable coexistence between strategies that would otherwise be outcompeted. These spatio-temporal structures may take the form of spiral waves, irregular waves, spatio-temporal chaos, frozen patchy patterns, and various geometrical configurations." Though Lindgren's model is not explicitly economic, the contrast he draws between an agent space in which interactions are structured by (relatively sparse) social networks and an agent space in which all interactions are possible (as is the case, at least in principle, with the impersonal markets featured in general equilibrium analysis) is suggestive. Padgett's paper offers a similar contrast, in a quite different context. Both Durlauf and Krugman explore the emergence of geographical segregation. In their models, agents may change location--that is, change their position in a social structure defined by neighbor ties. In these models (especially Durlauf's), there are many types of agents, and the question is under what circumstances, and through what mechanisms, do aggregate-level "neighborhoods" arise, each consisting predominantly (or even exclusively) of one agent type. Thus, agent's choices, conditioned by current network structure (the agent's neighbors and the neighbors at the sites to which the agent can move), changes that structure; over time, from the changing local network structure, an aggregate-level pattern of segregated neighborhoods emerges. No Global Controller No global entity controls interactions. Instead, controls are provided by mechanisms of competition and coordination between agents. Economic actions are mediated by
84 85 legal institutions, assigned roles, and shifting associations. Nor is there a universal competitor--a single agent that can exploit all opportunities in the economy Cross-cutting Hierarchical Organization The economy has many levels of organization and interaction. Units at any given level behaviors, actions, strategies, products typically serve as `building blocks' for constructing units at the next higher level. The overall organization is more than hierarchical, with many sorts of tangling interactions (associations, channels of communication) across levels Continual Adaptation Behaviors, actions, strategies, and products are revised continually as the individual agents accumulate experience--the system constantly adapts Perpetual Novelty Niches are continually created by new markets, new technologies, new behaviors, new institutions. The very act of filling a niche may provide new niches. The result is ongoing, perpetual novelty Out-of-Equilibrium Dynamics Because new niches, new potentials, new possibilities, are continually created, the economy operates far from any optimum or global equilibrium. Improvements are always possible and indeed occur regularly. Systems with these properties have come to be called adaptive nonlinear networks. (The term is John Holland's, 1987.) There are many such in nature and society: nervous systems, immune systems, ecologies, as well as economies. An essential element of adaptive nonlinear networks is that they do not act simply in terms of stimulus and response. Instead they anticipate. In particular, economic agents form expectations--they build up models of the economy and act on the basis of predictions generated by these models. These anticipative models need neither be explicit, nor coherent, nor mutually consistent. Because of the difficulties outlined above, the mathematical tools economists customarily use, which exploit linearity, fixed points, and systems of differential equations, cannot provide a deep understanding of adaptive nonlinear networks. Instead, what is needed is new classes of combinatorial mathematics and population-level stochastic processes, in conjunction with computer modeling. These mathematical and computational techniques are in their infancy. But they emphasize the discovery of structure and the processes through which structure emerges across different levels of organization. This conception of the economy as an adaptive nonlinear network--an evolving, complex system--has profound implications for the foundations of economic theory and for the way in which theoretical problems are cast and solved. We interpret these implications as follows: Cognitive foundations Neoclassical economic theory has a unitary cognitive foundation: economic agents are rational optimizers. This means that (in the usual interpretation) agents evaluate uncertainty probabilistically, revise their evaluations in the light of new information via Bayesian updating, and choose the course of action that maximizes their expected utility. As glosses on this unitary foundation, agents are generally assumed to have common knowledge about each other and rational expectations about the world they inhabit (and of course co-create). In contrast, the Santa Fe viewpoint is pluralistic. Following modern cognitive theory, we posit no single, dominant mode of cognitive processing. Rather, we see agents as having to cognitively structure the problems they face--as having to "make sense" of their problems--as much as solve them. And they have to do this with cognitive resources that are limited. To "make sense," to learn, and to adapt, agents use variety of distributed cognitive processes. The very categories agents use to convert information about the world into action emerge from experience, and these categories or cognitive props need not fit together coherently in order to generate effective actions. Agents therefore inhabit a world that they must cognitively interpret--one that is complicated by the presence and actions of other agents and that is ever changing. It follows that agents generally do not optimize in the
85 86 standard sense, not because they are constrained by finite memory or processing capability, but because the very concept of an optimal course of action often cannot be defined. It further follows that the deductive rationality of neoclassical economic agents occupies at best a marginal position in guiding effective action in the world. And it follows that any "common knowledge" agents might have about one another must be attained from concrete, specified cognitive processes operating on experiences obtained through concrete interactions. Common knowledge cannot simply be assumed into existence. Structural foundations In general equilibrium analysis, agents do not interact with one another directly, but only through impersonal markets. By contrast in game theory, all players interact with all other players, with outcomes specified by the game's payoff matrix. So interaction structures are simple and often extreme--one-with-all or all-with-all. Moreover, the internal structure of the agents themselves is abstracted away.[4] In contrast, from a complexity perspective, structure matters. First, network-based structures become important. All economic action involves interactions among agents, so economic functionality is both constrained and carried by networks defined by recurring patterns of interaction among agents. These network structures are characterized by relatively sparse ties. Second, economic action is structured by emergent social roles and by socially-supported procedures--that is, by institutions. Third, economic entities have a recursive structure: they are themselves comprised of entities. The resulting "level" structure of entities and their associated action processes is not strictly hierarchical, in that component entities may be part of more than one higher-level entity and entities at multiple levels of organization may interact. Thus reciprocal causation operates between different levels of organization--while action processes at a given level of organization may sometimes by viewed as autonomous, they are nonetheless constrained by action patterns and entity structures at other levels. And they may even give rise to new patterns and entities at both higher and lower levels. From the Santa Fe perspective, the fundamental principle of organization is the idea of that units at one level combine to produce units at the next higher level.[5]
What counts as a problem and as a solution
2. Norman Packard's contribution to the 1987 meeting addresses just this problem with respect to the dynamical systems approach. As he points out, "if the set of relevant variables changes with time, then the state space is itself changing with time, which is not commensurate with a conventional dynamical systems model."
3. John Holland's outline at the 1987 meeting beautifully--and presciently--frames these features. For an early description of the Santa Fe approach, see also the program's 1989 newsletter, "Emergent Structures."
4. Except in principal-agent theory or transaction-costs economics, where a simple hierarchical structure is supposed to obtain.
5. We need not commit ourselves to what constitutes economic "units" and "levels". This will vary from problem context to problem context.
6. There is a voluminous sociological literature on interaction networks. Recent entry points include Noria and Eccles (1992), particularly the essay by Granovetter entitled "Problems of Explanation in Economic Sociology", and the methodological survey of Wasserman and Faust (1994).
Last Modified: Monday, December 17, 2001
86 87
The End of Certainty in Economics
W. Brian Arthur
Talk delivered at the Conference Einstein Meets Magritte, Free University of Brussels, 1994. Appeared in Einstein Meets Magritte, D. Aerts, J. Broekaert, E. Mathijs, eds. 1999, Kluwer Academic Publishers, Holland. Reprinted in The Biology of Business, J.H. Clippinger, ed., 1999, Jossey-Bass Publishers
The story of the sciences in the twentieth Century is one of a steady loss of certainty. Much of what was real and machine-like and objective and determinate at the start of the century, by mid-century was a phantom, unpredictable, subjective and indeterminate. What had defined science at the start of the century—its power to predict, its clear subject/object distinction—no longer defined it at the end. In the century just past, science after science lost its innocence. Science after science grew up.
What then of economics? Is economics a science? I believe it is. It is a body of well-reasoned knowledge. Yet until the last few years it has maintained its certainty, it has escaped any loss of innocence. And so we must ask: Is its object of study, the economy, inherently free of uncertainties and indeterminacies? Or is economics in the process of losing its innocence and thereby joining the other sciences of this century? I believe the latter. In fact, there are indications everywhere these days in economics that the discipline is losing its rigid sense of determinism, that the long dominance of positivist thinking is weakening, and that economics is opening itself to a less mechanistic, more organic approach. In this chapter I will show my own version of this loss of certainty. I will argue that there are major pockets of uncertainty in the economy. I will show that the clear subject/object distinction in the economics often blurs. I will show that the economy is not a gigantic machine, but a construct of its agents. These are not “anomalies” to be feared, they are natural properties of the economy, and if we accept them, we will have a stronger, not a weaker science.
High Modern Economics The fundamental ideas in economics stem from the thinking of the eighteenth century, in particular from the thinking of the English and Scottish Enlightenment. In 1733, at the height of the intoxication of enlightenment thinking, Alexander Pope condensed its essence in one stanza of his poem, An Essay on Man: All Nature is but Art unknown to Thee All Chance, Direction, which thou canst not see All Discord, Harmony, not understood All partial Evil, universal Good: And, spite of Pride, in erring Reason’s spite One truth is clear, “Whatever IS, is RIGHT.” In this context “Art” means artifice. It means technique or mechanism. And so, all the intricate wonders we see in nature, says Pope, are in fact a gigantic machine, an artifice like the mechanical automata figures of his time. All that looks unkiltered really has direction behind it. All that looks complex and discordant, like the movements of planets before Kepler’s and Newton’s times, has a hidden simplicity. All that affects each of God’s creations adversely, in some unspoken way works to the good of the whole. Quoting Socrates, “Whatever is, is right.” These were not merely the ideas of Pope. They were the ideas that filled the intellectual air when Adam Smith was growing up. And Smith went on to enshrine them in The Wealth of Nations, that magnificent work that uncovered the
87 88 hidden simplicity behind the traffickings of traders and manufactories and butchers and bakers. The economy was indeed Art, and its principles were now unhidden. The selfish interests of the individual were guided as by an invisible hand to the common interest of all. Whatever was, was right. Two centuries later, the philosopher of science, Jacob Bronowski, was to comment glumly that economics never recovered from the fatally rational structure imposed on it in the eighteenth century. But we inherited more than Smith’s rational structure. Deep in some recess of our minds, we inherited the thinking that the economy is but Art, a gigantic machine, that if we merely understood its parts, we could predict the whole. Certainly when I was studying economics in Berkeley 25 years ago, many economists hoped (as I did) that a Grand Unified Theory of economics was possible. From the axioms of rational human behavior, a theory of the consumer could be constructed. From this and a corresponding theory of the firm we could construct a consistent microeconomics. From this, somehow, we could construct an aggregate theory of the economy: macroeconomics. All this would constitute a Grand Unified Theory of the economy. There have always been two embarrassments to this hope of constructing a theory of the economy from its reductionist parts. One was that the economy relies on human beings, not on orderly machine components. Human beings with all their caprices and emotions and foibles. The second embarrassment was technology. Technology destroys the neatness because it keeps the economy changing. Human behavior was finessed in economics by the device of Economic Man, that perfectly rational being who reasons perfectly deductively on well-defined problems. And technology change was not so much finessed as ignored, or treated as exogenous. And so to make an orderly, predictive theory possible, Economic Man (the subject) needs to operate on well-defined Problems (the object). There should be no blurring of agent and problem. And the well-defined problems should have well-defined Solutions. And the solutions would comprise the building blocks for the next aggregated level of the theory. This approach works. But it runs into difficulties when problems start to involve more than one decision maker and any degree of complication. Then heroic assumptions must be made. Otherwise well-definedness unravels, agent and problem become blurred, and pockets of uncertainty start to bulge. Let me illustrate what I mean in the context of a typical microeconomic situation in modern economics. (I have chosen it from the mid-1980s literature on industrial organization.) Consider this problem: We have a circle that we might think of as a 24-hour clock. A number of firms, say twenty airline companies, have to decide in which time slot of this clock their planes will take off, say from La Guardia Airport to go to Washington. Of course the different airlines have different preferences about when to take off. They know their preferences and are going to book such take-off slots. The choices will be made once and for all. But there is a trade off (in every decent economic problem there is always a trade-off) between where they really want to take off versus not being too close to other airlines’ choices of their time slots. So, given the airlines preferences, which time slots will they choose? This is the problem. We might feel uneasy about saying much with certainty here. But I want to show the modern version of the Enlightenment approach, where we find the Harmony of a solution within the Discord of the situation. This High Modern approach is called rational expectations. I will first spell it out, then shine a bright light of realism on it, so that it starts to unravel and pockets of uncertainty appear. Let’s go ahead. In the modern approach, we begin by supposing we know the order in which the airlines will submit their choices. Now imagine airline number 20, the last to choose, reasons like this: knowing where the first nineteen airlines are, I will know where I will want to be. So regardless of any arbitrary choice of the first nineteen airlines, I will know which time-slot to choose. This is an easy problem for me as the twentieth. What about airline number 19? Well, airline number nineteen, when choosing, will know the chosen positions of the previous eighteen airlines and can figure what it should do, given that the twentieth will choose an optimal position given the positions of the eighteen other airlines and number 19’s choice. What about the number 18? Well, the eighteenth, knowing what the previous seventeen have chosen, arbitrarily can solve the problem of selecting an optimal placement knowing what the nineteenth will do, given that the nineteenth makes his optimal choice, given what the twentieth will do as a result of number 19’s choice. Getting complicated? Yes. But you can work the whole logic in reverse order by backward deduction, or more properly by dynamic programming, and deduce how all twenty airlines will place themselves. Notice the properties of this procedure: The problem is well defined by making it sequential and assuming the airlines use logical backward deduction. The solution is precise and clean in a mathematical sense. The problem becomes a mathematical one. (Indeed all such problems become mathematical. And economics in turn becomes mathematics.) Another property that we normally have in this kind of problem is that the individual act comes to good of the whole, that is, partial evil is universal good. It is not quite true in this case, but nevertheless this is a generic
88 89 property that often holds in economics. However, the Solution comes with a lot of fine print. Airlines must know their preferences exactly. Not only that, they must know the preferences of all other airlines. Further they must know that every other airline accurately know the preferences of every other airline. They also must know that every airline knows that every airline knows the preferences of every other airline, and so on in an infinite regress. Also, each airline must be rational enough to work out the solution. Further, each airline must believe that every other airline is rational and will use perfect rationality to work out the solution. Further each airline must know in an infinite regress that every other airline is using this rational way to work out the problem, because if one of these airlines fails to do so, it messes the solution up for every other airline. Further the optimal placement of each airline using this backward deduction must be unique. If any link of this network of requirements breaks, the solution ceases to exist. This type of multi-agent choice problem is pervasive in economics. So let us take this solution approach seriously. What if we are airline number 3 and we feel uncertain as to what airline number 17 is going to do? As airline number 3, we might say: “I don't think the people of airline number 17 are that bright, and I'm not sure whether they are going to solve this problem by this rational method. And if they don't work it out in this way then I am not sure what my optimal choice would be as the third bidder in the process.” This is sufficient to upset the situation. But worse, airline number 3 may communicate its uncertainty to other airlines and they may no longer rely on number 3 or number 17. The entire solution is starting to unravel. In fact the Solution as defined by rational expectations theory is a function of airlines’ expectations or predictions of what other airlines are going to do. The problem is that if I am a representative airline I am trying to figure out what my expectations ought to be—I am trying to predict a world that is created by the expectations of myself and everybody else. There is a self-referential loop here. The outcome each airline is trying to predict depends on the predictions it and others might form. In other words, predictions are forming a world those predictions are trying to forecast. Barring some coordinating device, by which an airline can logically determine the predictions of others (such as the tortured solution-reasoning above), there is no logical way it can determine its prediction. There is a logical indeterminacy. So it is in the economy. People are creating a world that forms from their predictions, but if they try to form these expectations in a perfectly logical deductive way, they get into a self-referential loop. There is a logical hole in standard economic thinking. Our forecasts co-create the world our forecasts are attempting to predict. And if I do not know how others might determine their forecasts, mine are indeterminate. There are some cases in economics where it is pretty obvious that everyone can figure out what to do, where something like the above given scheme does work. But otherwise the problem is fundamental. When our ideas and preferences co-create the world they are trying to forecast, self-reference renders the problem indeterminate. The idea that we can separate the subjects of the economy—the agents who form it—from the object, the economy itself, is in trouble. Pockets of indeterminism are present everywhere in the economy. And the high modern form of economic determinism fails.
Economics under Indeterminacy There are two questions we want to ask now. One question is: Does it matter ? Maybe all of this happens on a set of measure zero, maybe this difficulty is confined to some trivial examples in economics. The second question is: If there are pockets of indeterminism how should we proceed? To answer these I want to turn to the field of capital markets, to asset pricing theory—an area of economics that does matter. There is a well worked out efficient-market economic theory for financial markets and there is a very different set of ideas that financial traders use. Let me first outline the standard theory. The standard efficient markets theory says that all and any information hinting about the future changes of the price will be used by investors. By an argument very much like the airline argument, each stock’s price is bid to a unique level that depends on the information currently available. Using past patterns of prices to forecast future prices (technical trading), in this view, cannot lead to further profits. Otherwise the information inherent in past prices could be used to make further profits, and by assumption investors have already discounted all useful information into current prices. So the standard theory says investors use all information available to form expectations. These will determine stocks’ prices which on average will uphold these same expectations. Rational expectations again. Thus there is no way to make any money, and the market is efficient. Traders, on the other hand, believe that the market is forecastible. They believe they can spot patterns in past prices helpful in prediction—they believe in technical trading. They believe the market is anthropomorphic, that it has a psychology, that it has motives. “The market was nervous yesterday. But it shrugged off the bad news and went on to quiet down.” Economists are skeptical of this, and so the two viewpoints sit badly with each other.
89 90
The standard theory is wonderfully successful. It has its own logic. And this logic is complete and has desirable properties such as uniqueness of solution. But the standard theory must face some unexplained phenomena— or so-called empirical anomalies. Crashes and bubbles seemingly with no cause. The fact that the volume of market trades is an order of magnitude higher than theory predicts. The fact that econometric tests show that that technical trading is indeed profitable statistically (Brock, Lakonishok, and Le Baron). The phenomenon of GARCH behavior, (GARCH means Generalized Auto Regressive Conditional Heteroscedasticity), which means there are periods of high volatility in stock prices interspersed randomly with periods of quiescence. In sum, the standard theory does not explain at least half a dozen major statistical “anomalies” in real markets. This has recently led to a great deal of modern thinking, some using ad-hoc behavioral observation, some more sophisticated theorizing. Let me now show, as in the airline problem, how the standard theory breaks down and leads to pockets of indeterminacy. Suppose investors can put some portion of their money in a single stock that pays a dividend every time period (a day, a year, say), and they cannot perfectly predict this dividend. The investors are buying the stock for the dividend plus any capital appreciation (tomorrow’s price), and they face the problem of forecasting these. To make the standard solution work, we assume homogeneous, identical investors who have identical forecasts of the dividend at the end of the period and identical forecasts about the stock’s price in the future—forecasts that are on average unbiased and are therefore rational expectations. A little economic reasoning then shows today’s price is equal to the common expectation of tomorrow’s price plus dividend (suitably discounted and weighted). This yields a sequence of equations at each time, and with a pinch or two of conditional-expectation algebra, we can solve these for the expectations of future prices conditioned on current information, and wind up with today’s price expressed as a function of expected future dividends. The problem is solved. But it is only solved, providing we assume “identical investors who have identical forecasts of the dividend at the end of the period and identical forecasts about the stock’s price in the future.” But what if we don’t? What if we assume investors differ? Let us look at the same exercise assuming our investors agents are not homogeneous. Note that the standard theory’s requirement of identical “information” means not just the same data seen by everyone, but the same interpretation of the data. But imagine yourself in a real financial market, like the New York stock market. Then this information consists of past prices and trading volumes, moves made by large mutual funds or large pension funds, rumors, CNN, network news, the market section of the Wall Street Journal, what other traders are doing, what they are telling you by telephone, what your friend’s uncle thinks what is happening to the market. All of these things compromise actual information and it is reasonable to assume that, even if everybody has identical access to all this information, they would treat this information as a Rorschach inkblot and would interpret it differently. Even if we assume that the people interpreting this information are intelligent to any arbitrarily high degree and they are all perfectly trained in statistics, they will still interpret this data differently because there are many different ways to interpret the same data. So there is no single expectational model. A given investor can still come up with an individual forecast of the dividend. But tomorrow’s price is determined by this investor’s and other investors’ individual forecasts of the dividend and of next period’s price. And there is no way for our reference investor to fathom the forecasts of the others—to figure “what average opinion expects the average opinion to be” (to use Keynes’ words). To do so brings on a logical regress. “I think that they might think such and such, but realizing that I think that, they will think this.” Unless we assume identical investors, once again our agents are trying to forecast an outcome (future price) that is a function of other agents’ forecasts. As before there is no deductive closure. Expectations become indeterminate, and the theory collapses. Worse, expectations become unstable. Imagine that a few people think that prices on the market are going to go up. If I believe this and I believe that others believe this, I will revise my expectations upward. But then I may pick up some negative rumor. I will reassess downward, but realizing that others may reassess and that they too realize that others, I may further reassess. Expectations become fugitive, rippling up or down whether trades are made or not. Predictions become unstable. This is the way price bubbles start. If somehow people expect prices to go up, they will forecast that other people will forecast that prices will go up. So they will buy in, and once the bubble thus starts off, people can see prices go up and their expectations of upward motion fulfilled. Therefore prices may continue to go up. Similar logic applies to “floors” and “ceilings.” If, for example, the price is 894, many investors believe that at 900 there is some sort of membrane, a ceiling, and when the price reaches this ceiling it will bounce back down with a certain probability or it may “break through.” Such ideas seem strange at first. But it is quite possible that many investors have sell orders at 900, simply because it is a round number. So expectations that the price will fall if it hits 900 are likely to be fulfilled.
90 91
Ceilings and floors emerge as partially self- fulfilling prophesies, held in place by their being convenient sell and buy places. We are now a long way from homogeneous rational expectations. Under the realistic assumption that traders may interpret the same information differently, expectations become indeterminate and unstable. And they may become mutually self-fulfilling. To summarize all this: If we look at a serious branch of economics, the theory of capital markets, we see the same indeterminacy that we saw in the airline problem. Agents need to form expectations of an outcome that is a function of these expectations. With reasonable heterogeneity of interpretation of “information,” there is no deductive closure. The formation of expectations is indeterminate. And yet in every market, in every day, people do form expectations. How do they do this? If they can not do this deductively, then is it possible to model their behavior in this area ? In 1988, John Holland and I decided that we would study situations like this by forming an artificial stock market in the computer and giving the little agents—artificially intelligent computer programs—some means by which they can do the reasoning that is required. This was one of the very earliest artificial, agent-based markets. Later we brought in Richard Palmer who is a physicist, Paul Tayler who is a finance expert and Blake LeBaron who is a financial theorist in economics. In this market there was no feed-in from the real stock market. It was an artificial world going on inside the machine. The artificial agents, the little artificial investors, are all buying and selling a “stock” from one another. The computer could display the stock’s price and dividend, who is buying and selling, who is making money and who is not, who is in the market and who is out, and so on. The price is formed within the machine by bids and offers. And another little program—a specialist—sets the price to clear the market, as in actual stock markets. The modeling question was: If the agents cannot form their expectations deductively, how are they going to form them? We decided to follow modern cognitive theory about how actual human beings behave in such situations. So we allowed our artificial agents looking at to posit multiple, individual hypothetical models for forecasting, and to test these on a continual, ongoing basis. Each of these hypotheses has a prediction associated with it. At any stage each agent uses the most accurate of its hypotheses, and buys or sells accordingly. Our agents learn in two ways: they learn which of their forecasting hypotheses are more accurate, and they continually toss out ones that do not work and replace these using a genetic algorithm. So they are learning to recognize patterns they are collectively creating, and this in turn collectively creates new patterns in the stock price, which they can form fresh hypotheses about. This kind of behavior—bringing in hypotheses, testing them, and occasionally replacing them—is called induction. Our agents use inductive rationality—a much more realistic form of behavior. Very well. But now the key question is: Does our market converge to the rational expectations equilibrium of the academic theory or does it show some other behavior? What we found to our surprise was that two different regimes emerged. One, which we called the rational expectations regime, held sway when we started our agents off with sets of predictive hypotheses close to rational expectations. We could plot the parameters of all the predictive hypotheses on a chart, and in this case, over time, we could watch them getting gravitationally pulled into the orbit of the rational expectations solution, forming a “fuzz” around this point, as they made occasional predictive forays away from rational expectations to test different ideas. It is not hard to see why rational expectations prevailed. If the overall mass of predictions is near rational expectations, the price sequence will be near rational expectations, and non-rational expectations forecasts will be negated. So the academic theory was validated. But there was a second regime, which we called the complex regime, and it prevailed in a much wider set of circumstances. We found that if we started our agents with hypotheses a little removed from rational expectations, or alternatively, if we allowed them to come up with hypotheses at a slightly faster rate then before, the behavior of the market changed. Subsets of mutually reinforcing predictions emerged. Imagine, for example, we have a 100 artificial agents each using 60 different prediction formulas, so that there is a universe of some 6,000 predictors. Some of the predictors that emerge are mutually reinforcing, some are mutually negating. Suppose many predictors arise that say the stock price cycles up and down over time. Such predictors would be mutually negating because they will cause agents to buy in at the bottom of the cycle, and sell at the top of the cycle, mutually negating profits, and therefore eventually disappearing from the population of predictors. But if a subset of predictors emerged by chance that said “the price will rise next period if it has risen in the last three periods,” and there were enough of these, they would cause agents to buy, which on average would cause the price to rise, reinforcing such a sub-population. Such subsets could then take off, and
91 92 become embedded in the population of predictors. This was what indeed happened in the complex regime, endowing it with much richer set of behaviors. Another way to express this is that our artificial traders discovered forms of technical trading that worked. They were using, with success, predictions based upon past price patterns. And so technical trading was emergent in our artificial stock market. This emergence of subsets of mutually reinforcing elements, strangely enough, is reminiscent of the origin of life, where the emergence of subpopulations of RNA in correct combinations allows them to become mutually enforcing. Another property that emerged in the complex regime was the so-called GARCH behavior I mentioned earlier that occurs in real markets—periods of high volatility in the stock price followed by periods of quiescence—which is unexplained in the standard model. How did GARCH become an emergent property? What happens in our artificial market is that every so often some number of investors discover a new way to do better in the market. These investors then change their buying and selling behavior. This causes the market to change, even if slightly, possibly causing other investors in turn to change. Avalanches of change sweep through the market, on all scales, large and small. Thus emerge periods of change triggering further change—periods of high volatility—followed by periods when little changes and little needs to be changed, periods of quiescence. This is GARCH behavior. Let me now summarize. What we found in our artificial stock market is that, providing our investors start near the academic rational-expectations solution, this solution prevails. But this is a small set of parameter space. Outside this, in the complex regime, self-reinforcing beliefs and selfreinforcing avalanches of change emerge. A wider theory and a richer “solution” or set of behaviors then appears, consonant with actual market behavior. The rational-expectations theory becomes a special case. In the standard view of the economy, which has an intellectual lineage that goes back to the enlightenment, the economy is mechanistic. It is complicated but can be viewed as a series of objects and linkages between them. Subject and object—agents and the economy they perform in—can be neatly separated. The view I am giving here is different. It says that the economy itself emerges from our subjective beliefs. These subjective beliefs, taken in aggregate, structure the micro economy. They give rise to the character of financial markets. They direct flows of capital and govern strategic behavior and negotiations. They are the DNA of the economy. These subjective beliefs are a-priori or deductively indeterminate in advance. They co-evolve, arise, decay, change, mutually reinforce, and mutually negate. Subject and object can not be neatly separated. And so the economy shows behavior that we can best describe as organic, rather than mechanistic. It is not a well-ordered, gigantic machine. It is organic. At all levels it contains pockets of indeterminacy. It emerges from subjectivity and falls back into subjectivity.
92 93
93 94
Is the Information Revolution Dead? If history is a guide, it is not.
By W. Brian Arthur, March 2002 Issue
At the peak of the Internet frenzy two years ago, when the Nasdaq was over 5,000 and dotcom millionaires were buying spreads in the hills above Palo Alto, it seemed that the information revolution would go on forever. Little tech companies were popping up everywhere, and small investors were reaping returns that made them feel like geniuses. Then the bubble burst. It burst, management guru Peter Drucker tells us, because "the information industry as a business wasn't going anywhere." The information revolution had been hyped, exaggerated. Neither computers nor the Internet, Drucker says, had added much to the economy.
Is the information economy going nowhere? Is its revolution over? In Silicon Valley, certainly, the prospects look bleak. But history suggests that such pessimism is misplaced -- that the information revolution's best days might actually lie ahead.
Join W. Brian Arthur and Andy Grove on March 18 in San Jose for a discussion of "Is the Information Revolution Dead." Register now at Business 2.0 Live!.
Looked at without historical context, the information revolution appears to be unique, comparable to nothing we know about from before. Looked at as part of history, it is merely one in a series of technological revolutions that have been occurring since the mid-18th century. Each of these revolutions has been different: The Industrial Revolution, from about 1760 to 1820 in Britain, replaced handcrafting with machinery and brought the factory and mill system. The railway revolution, from about 1825 to 1875, again in Britain, saw a great connecting of commerce and the coming of steam power. The steel and electricity revolution, from about 1875 to 1920 (the action now shifts to the United States and Germany), was an age of massive engineering and the electrification of the economy. The great manufacturing age, from 1910 to 1970 or so, brought us mass production and automobiles and cheap goods aplenty. And our own revolution, which started with the microprocessor in about 1970, brings us the age of digital everything, the Web, and interconnected commerce. The dates I've given are approximate. Economists quarrel over when such eras started and ended, and about which clusters qualify as "revolutions." Some deny that "eras" of great change exist at all. Others, notably economists such as Carlota Perez and Chris Freeman of the Sussex school in England, champion the notion of revolutions and see in their phases portents of what is to come.
94 95
All threads of thought on technology revolutions lead back to Austrian economist Joseph Schumpeter, a single figure writing in the first half of the 20th century. Schumpeter has a curious position in economics. He is revered on the continent of Europe, yet has a shadowy reputation in Anglo-Saxon economic circles -- you can get a graduate degree in economics at an English or American university and scarcely hear of him. He is remembered more by business gurus for his idea of innovation bringing "gales of creative destruction." Professors who do speak of him are fond of telling their classes that Schumpeter aspired early in life to be the greatest economist in the world, the greatest horseman in Austria, and the greatest lover in Vienna. The story smacks of myth. But as far as I can track it down, it is true, and late in life Schumpeter is said to have only admitted that he was not the greatest horseman in Austria. Was he the greatest economist in the world? He was certainly not considered so during his lifetime -- others, his nemesis John Maynard Keynes among them, were better known. But I believe that Schumpeter will turn out to be the most important economist of the 20th century. He concerned himself not with an economy at rest but with the unfolding of economies, with their ongoing tendency to evolve and develop and change in structure. And this he ascribed to innovation -- to ongoing, disruptive discoveries in technology and their incorporation in the economy. His writings, some now nearly 100 years old, are surprisingly modern.
Schumpeter noticed that technology arrives in clusters -- with electrification come dynamos, generators, transformers, switch gear, power distribution systems; with mass production and the automobile come production lines, modern assembly methods, "scientific management," road systems, oil refineries, traffic control. These clusters, if they are important, define an era. They eventually change the way business is done, even the way society is conducted. As Perez tells it, a technology revolution starts with the opening up of one or more technologies that "enable" the new cluster. The new technology cluster, at first little noticed, achieves successes in early demonstrations, and technical people start small companies based on the new ideas. These compete intensely in this early turbulent phase. Government regulation is largely absent, and as successes mount in a technical free-for-all, the promise of extraordinary profit looms. The public starts to speculate. (In the mass-production revolution, think of the 1920s in the United States.) The middle phase sees a sustained buildout or golden age of the technology, during which it becomes the engine of growth for the economy. Large companies and oligopolies reign, and the period is one of confidence and prosperity. (Think of the 1950s and '60s.) In the last phase, the technology is mature. It has saturated its possibilities, production moves to places on the periphery, and complacency sets in. (Think of the 1970s and the rise of competition in Japan and Taiwan.) Profits at home are low, and entrepreneurs begin to look around for new opportunities. The economy becomes ripe for the next revolution.
The exact phases and what happens within them are debatable. But what interests me is the pattern of speculative exuberance, followed by crash, followed by a strong buildout period. If the Schumpeter-Perez-Freeman story holds water, we are not at the end of the information revolution. We are only partway into it, and the buildout -- the golden age -- has yet to come.
If we lay the information revolution alongside the great railway revolution in Britain, year for year, we'd now be somewhere around 1850 -- just after the railway investment mania of 1845 and its crash in 1847. The railway revolution took place roughly between 1825 and 1875. I say roughly, because there never is a marked beginning or end to an economic revolution. Even in 1825, railways were by no means new. For centuries, mines had used horse-drawn wagons to move ores on wood or iron-capped rails. The first commercial railway, the Stockton & Darlington, owned a single locomotive when it opened in 1825, and its Express was still a carriage drawn on rails by a cantering horse. Even when the Liverpool & Manchester Railway was being planned in 1829, its
95 96 directors doubted that a moving locomotive could retain adhesion on uphill gradients. The conventional view bet on carriages hauled by ropes attached to stationary winding engines. In October 1829, the L&M organized a locomotive trial at Rainhill, stipulating that the engine must be capable of "drawing a train of carriages at 10 mph." Five locomotives entered, and Robert Stephenson's Rocket astonished the watching gentlemen by achieving 24 mph unloaded and 12 mph hauling a train up the Rainhill incline. Locomotives had proved themselves. Technical pioneers began to crowd in, a host of improvements followed, and a decade and a half of frenzied technical competition was under way.
The new technology engendered talk of a new, more prosperous economy. An 1831 prospectus for the London & Essex Railroad promised that "the first necessaries of life will be supplied in greater abundance; competition increased, and a reduction in prices the necessary consequence." Railways became fashionable. Queen Victoria made her first railway trip in 1842 in a suitably imperial carriage and allowed that she was "quite charmed by it." Entrepreneurs began to emerge, among them George Hudson, the "railway king." Hudson started as a draper in York, inherited money, and found he had a talent for putting together new and branch lines. He became a public figure, fawned over, known as His Steam Majesty. A contemporary observer, John Francis, recorded that "his fortune was computed with an almost personal pride.... The choicest aristocracy ... sought his presence.... The bishop bent in homage.... When his name graced an advertisement, men ran to buy the shares. He was their railway potentate; their iron king; their golden god." The railway kings such as Hudson were not so much technical people as organizers and investors -- mainly of other people's money. And indeed they organized and invested. And got rich. They bought large estates -- Hudson purchased the magnificent 100-acre Newby Park estate from Earl de Grey in 1845.
Hudson was one of many railway entrepreneurs. The procedure in proposing a railway required bringing a bill before Parliament; if it was approved, the stock could be subscribed in a sort of Victorian IPO. Gentlemen and politicians and dignitaries such as the Duke of Wellington bought stock. Those who could not afford stock could buy scrip -- shares diced and parceled into small units that could be sold on the street. Servants and spinsters and tradesmen began to pour their savings into railway scrip.
By 1845 a full railway mania was raging. By the summer new schemes were being floated at the rate of more than a dozen a week. Scrip was sold by alley men, and the stock exchange resembled a country fair. The general prosperity contributed. "The markets were good ... and all was smiling," Francis wrote. "The most cautious were deceived by this apparent prosperity.... Like drunken men they lost their caution and gave their signatures to everything that was offered.... Many of the railways attained prices which staggered reasonable men. The more worthless the article, the greater seemed the struggle to attain it." Schemes for direct lines connecting little-known towns to other little-known towns became a craze, launched more with an eye to garnering investment than actual profits. "The country," said Lord Cockburn, a Scottish judge, "is an asylum of railway lunatics." Not all schemes could be profitable. "We see nine or ten proposals for nearly the same line, all at a premium, when it is well known that only one CAN succeed," said the Economist. The result, predictably, was overcapacity. There were three independent routes from London to Peterborough, and three from Liverpool to Leeds.
Trouble began in October 1845, when scrip ceased to pay a premium and shares in established railways began to fall. Great Western Railway shares had plummeted 40 percent from their August peak. A harvest failure in 1846 compounded the downswing, further bankruptcies followed, and an economic Week of Terror began on Monday, Oct. 17, 1847. Some banks were
96 97 forced to close. The Bank of England held less than £2 million in reserves, and special measures had to be taken to stave off national economic collapse. When the panic was over, railway shares had lost 85 percent of their peak value and several hundred companies had folded. As Francis recorded it, "Entire families were ruined. There was scarcely an important town in England but what beheld some wretched suicide." Hudson himself was never convicted of wrongdoing. ("He was obviously no adept at the higher arts of swindling," Schumpeter remarked.) But he was now a pariah and fled to Paris. Thomas Carlyle wrote in 1850 of imagining Hudson swinging on a gibbet "as a tragic pendulum ... veritably the Supreme Scoundrel of the Commonwealth, who in his insatiable greed and bottomless atrocity had ... led multitudes to go, in the ways of gilded human baseness; seeking temporary profit ... where only eternal loss was possible." Many of Hudson's investors cherished a similar hope.
Fifty years earlier a similar story had played out in the canal mania of the 1790s. Canals had been around for a couple of hundred years, but got under way seriously in Britain in 1761 when the Duke of Bridgewater drove a canal from his coal mines at Worsley to the textile mills at Runcorn. In the two decades that followed, large profits from this and similar undertakings brought about the swift expansion of canal systems -- and a speculative mania in 1792. Canal shares crashed in 1793, to the ruin of many.
What is interesting about both the canal and railway revolutions is that their crashes were by no means the end. In the decades after 1793, Britain went on to build out 2,000 miles of waterway, doubling its precrash mileage. And canals became the key infrastructure component of the Industrial Revolution. Similarly, in 1845, just before the crash, Britain possessed 2,148 miles of railway; 65 years later it had 21,000 miles. The major buildout of railways came after the crash of 1847.
Of course, after a crash much of the glamour of the new technology is lost and is not easily replaced. The new period is different. The mood is different. If the period before 1847 is a time of excitement and of small companies jostling for dominance, the years after 1847 are ones of seriousness and hard work -- years of buildout rather than novelty, years of confidence and steady growth, years of orderliness. Investment profits begin to reflect the real returns from the new technology. The base technology is now in place, but before the crash it had not accounted for much in economic terms. In the decades following the 1847 crash, the railways come into their own. Passenger and freight receipts become multiples of what they were before, and the very growth of the railways helps the economy to grow, which further stimulates the railways. After 1850, railways become the engine of the economy in Britain.
In the United States, there was no equivalent of the British railway mania. Certainly there were periods of setback in which railroad overinvestment was partly to blame. In the depression of 1859, the economic commentator Henry Carey Baird complained that "our railroad system has cost more than $1,000,000 and has brought ruin upon nearly everyone connected with it, the nation included." But, again, at this time railroads in the United States were just beginning. In 1860 the United States had 30,000 miles of built-out track; by 1914 it had 253,000 miles. The buildout, when it came, was massive. And it brought an age of oligopolies and railroad barons. Yet in spite of such excesses, railroads became the driving force of the U.S. economy. Railroads opened up the West, they provided demand and thus the base for the new steel industry, they made possible new commerce and other new industries, they brought new cities and centers of population into existence, and as in Britain the commerce of the country organized itself around them.
Not all technology revolutions of the past exhibit manias and crashes. Economists dispute to what degree the great crash of 1929 was the result of overexuberance about the stocks of the new
97 98 automobile manufacturers and other mass producers. And there was no steel crash in the 1890s. But this dog-that-didn't-bark clue -- the absence of a steel crash and dubiousness of a mass- production crash -- doesn't negate any correlation between the railway crash and the Internet crash. It points to a resemblance that might otherwise be missed. Railways and canals, like the Internet, are connection technologies. They connect places, they connect businesses. As such they are natural monopolies -- only one line, or one canal, can profitably connect Liverpool to Manchester, and once this is put in place, competing lines lose. For connection technologies, this brings on a "race for space." And this in turn means that when the opportunities open up, they open not with an orderly funding but with a heated sense that they are finite and will soon be filled. The result can easily be an investment frenzy -- a mania. By contrast, steel factories can constantly improve and undercut one another, so that as the technology improves in an orderly way, it becomes financed in an orderly way. No mania. But with or without manias, all revolutions still progress from early chaotic innovation to buildout, and then to tired overcapacity and foreign competition.
The current technology cluster -- microprocessors, telecommunications, software, the Internet -- has had its boom and crash. And as with the railways, the telecommunications side of it suffers from overcapacity. It has made possible a host of subtechnologies and digitally based activities -- genomics, CT scanners, DNA probes, global positioning systems, cell phones -- and nobody doubts that it will spawn others. But the Internet part of this revolution seems to be in trouble and its prospects are derided. Is its buildout yet to come? Certainly we can match the Internet with what happened in the 1840s: the initial successes, the ballyhoo, the IPO kings, the large estates. And the crash. But an eerie resemblance to events of 160 years ago doesn't guarantee that the information revolution is about to enter any golden age. What evidence is there for this?
One argument I don't buy is the proposal that the information revolution has been around long enough to have had its chance. Its enabling technologies -- the laser, the microchip, the Arpanet (forerunner of the Internet) -- date from the 1950s to the early 1970s. So the revolution, this argument goes, has had 30 years to prove itself and we have yet to see its economic gains. Here again, history tells us otherwise. It tells us that a considerable delay -- several decades, usually -- lies between the technologies that set a revolution in motion and the revolution's heyday. The enabling technologies of the steel and electricity revolution (the Bessemer steel plant and the electric motor and generator) had arrived by the 1870s, but their full effects were not felt until well into the 1910s. Watt's steam engine was developed in the 1760s; steam power did not come into prevalent use until the 1820s. Modern mass production arrived in 1913, but did not reach its peak until the 1950s and '60s. Such decades-long delayed reactions should give us a clue that something other than the appearance of a technology and its subsequent adoption is at work. If a powerful new technology appears, it might take people a decade to hear about it and try it. But three decades? Five decades? Something else besides slowness to glom onto the new technology must be going on.
That something else, I believe, is that many arrangements, many improvements, and many organizational changes need to be put in place before the new technology cluster can become widespread. It is not enough that the base technologies of a revolution become available. Whether these are railroads or microchips, a revolution doesn't fully arrive until we structure our activities -- our organizations and business methods -- around its technologies, and until these technologies adapt themselves to us by becoming comfortable and easy to use. So it's not merely that the base technologies have to become better, faster, cheaper. That helps, but what's needed for the revolution to fully blossom are the 1,001 subtechnologies, arrangements, and architectures that adapt us to the new technologies and them to us. Their arrival takes time, and it defines the buildout period as one that creates the arrangements and subtechnologies that bring the new possibilities into full use.
98 99
We can see this with railways after the crash of 1847. Not only did rail transport become better, faster, cheaper -- with improvements such as steel rails arriving in the 1860s and, a decade later, compound locomotives that increased power by expanding steam in one cylinder after another. But systems that made railways usable and safe -- what I will call arrangements-of-use -- also followed in this period: lever systems that worked switches and signals, control of traffic via the electric telegraph, air brakes, double tracks, Pullman sleeping cars, dining cars, toilets. (Toilets arrived later on British trains than on American ones due to a certain English indisposition to admit bodily functions in public.) In the buildout period, innovation continues adapting the new technology to human use without letup. And business in turn adapts itself to the railway. Factories that might formerly have been located near rivers are now for convenience located near railways. Production facilities and shipping methods are adapted to the new transportation.
In fact, with any important revolution, business organization needs to do more than adapt. It needs, to some considerable degree, to redefine itself -- to re-architect itself. In 1955, economist Marvin Frankel noted that England's Lancashire textile mills still used old, out-of-date machinery. Yet the mill owners were aware of the advantages of modern methods and machinery. Why didn't they adopt them? Frankel found that the new, more efficient machines were heavier and larger, and the old Victorian brick mill buildings with their close supporting columns could not physically accommodate them or bear their weight. The new technology required tearing down the old structures and building different ones. The costs of this were prohibitive. Economic historian Paul David tells a similar story about the slow adoption of the electric motor. Before electrification, every factory was powered by a single steam engine, a giant hissing and cranking contraption with pistons and a flywheel and a system of belts and pulleys that turned shafts on several floors of a building to power all of the factory's machinery. Electric motors powered machines separately, and therefore often required factories to be redesigned. Industrial architects knew nothing of electricity, and finding the proper layout and organization took much experimentation. Full adoption took 40 years.
We are in a similar position today with computation and the Internet. Businesses routinely install digitally based equipment or enterprise software or peer-to-peer communications. But for effective use, they need to restructure their activities -- their very organization. How to do this, exactly, is not yet known. Businesses don't know what the appropriate organization will look like, and they face the possibility that any reorganization may well be quickly outdated by new technologies. Even if the correct organization were known, businesses would still have to pay the Marvin Frankel costs of tearing down the old structures. Much of the technology is available, but its use requires a slow process of learning and investment. In fact there is strong evidence that just such a process of tearing down old structures and steadily adopting information technology was under way in most U.S. industries all through the 1990s and that it statistically accounted for virtually all of the considerable productivity gains of that decade. Two recent studies show that this process continues -- that even after the recent crash, productivity gains issuing from the new technology have barely slackened.The reason is simple: Crash or not, information technology prices fall constantly, and businesses take advantage of this to purchase the new technology and adapt their methods to it.
This is healthy. But it is not sufficient that businesses and people adapt to a new cluster of technology. The real gains come when the new technology adapts to them. The notion that a technology needs to adapt to its users seems obvious enough, but is heavily underestimated. People will not use a technology that doesn't work properly. They will shun anything awkward or untrustworthy or just plain difficult to use. Making the technology better, faster, cheaper is only part
99 100 of what's needed. A new technology is used when it is more convenient, easier, reliable. For widespread use, a technology must provide, in a word, amenity.
As technologist John Seely Brown points out, something subtle happens to a technology when it achieves amenity: It disappears. We become absorbed into its world, and its bones don't stick out anymore. Thus, if we are driving a car at night, we are absorbed into car-world -- we are aware we are driving, aware of the passing trees and fences, but not aware of the car as a technology. The car disappears. By this standard the Internet is somewhere around the get-out-and- get-under days of the Model T Ford. To access my bank account, I have to fire up my computer, and wait. Then dial in by modem, and wait. Then get a browser going, and wait. Then enter account numbers and passwords, and wait. All the time there exists a barely noticeable anxiousness that the process may hang up at any moment. The Web's interface remains uncomfortable to use: ill-fitted, unreliable, frustrating, slow, and lacking content when we get there. It has not disappeared, nor has it achieved amenity. It will take time for amenity technologies to become available and used. It took automobiles from about 1890 to the 1940s -- half a century of development -- to reach amenity. Needed were paved roads, reliable brakes, ignition systems, safe tires, and a thousand other things.
The information revolution is not radically different from previous revolutions. The Internet has had its boom and crash, and there is no reason to suppose that history will be negated: Full use of the technology will arrive eventually. It always has. But this will require that the technology become workable for the user, and that businesses re-architect themselves to make use of it. This will happen gradually during the next 10 to 20 years as the missing components of the technology's use structure are put in place. In this buildout, the technologies that will matter most, that will determine the pace, are the ones I am calling arrangements-of-use. If there is one difference with this revolution, however, it is that it won't end when we have blanketed the country with optical cable or have teraflop processors. Information technology morphs every 10 years or so, so that what we thought defined the information revolution -- batch processing, desktop computing, Web-based interconnection -- is continually superceded by something new. What lies ahead can never be fully foreseen. This means that we can expect more innovation in this buildout phase than with previous revolutions. But during the next few years, at least, what will drive the buildout is something at once silent and unremarkable: the quiet, inexorable interconnection of business and the slow appearance of Web-based services that digitization provides.
How fast can the information technology economy come back? I don't know. The economy is quiet now, gestating a new phase. What I do know is that when that new phase comes forth, it will be a giant.
Further Reading
Carlota Perez, Technological Revolutions and Financial Capital, Edward Elgar Publishing, Cheltenham, U.K., forthcoming 2002.
Chris Freeman and Francisco Louca, As Time Goes By, From the Industrial Revolutions to the Information Revolution, Oxford University Press, 2001.
100 101
7.1. "Competing Technologies, Increasing Returns and Lock- in by Historical Events," Economic Journal, 99, 106-131,1989. [18 páginas]
Brian arthur EJ.pdf 7.2. "Positive Feedbacks in the Economy," Scientific American, Feb. 1990. [12 páginas]
brian arthur SciAm_Article.pdf 7.3. "Bounded Rationality and Inductive Behavior (the El Farol Problem), American Economic Review, 84,406-411, 1994. [11 páginas]
brian arthur El_Farol.pdf 7.4. Preface to the book: Increasing Returns and Path Dependence in the Economy, Univ. of Michigan Press, Ann Arbor, 1994. [9 páginas]
brian arthur Preface.pdf 7.5. "Complexity in Economic and Financial Markets," Complexity, 1, 20- 25, 1995. [14 páginas]
brian arthur Complexity_Jnl.pdf 7.6. "Increasing Returns and the New World of Business,"Harvard Business Review, July-Aug 1996. [10 páginas] vid p. 68 supra.
brian arthur HBR.pdf
7.7. "Process and Emergence in the Economy," introduction to the book The Economy as an Evolving Complex System II, edited by Arthur, Durlauf, and Lane, Addison Wesley, Reading, Mass, 1997. [14 páginas] vid p. 78 supra. 7.8. "The Economy as an Evolving Complex System II.," W. Brian Arthur, Steven N. Durlauf, and David A. Lane, (Eds.), Proceedings Volume XXVII, Santa Fe Institute Studies in the Science of Complexity, Reading, MA: Addison-Wesley, 1997. Review by Gerald Silverberg, Maastricht. [6 páginas]
101 102
brian arthur Silverberg_Web.pdf 7.9. W. Brian Arthur, John H. Holland, Blake LeBaron, Richard Palmer, and Paul Tayler, “Asset Pricing under Endogenous Expectations in an Artificial Stock Market”, December 1996 preprint. Final version published as pages 15--44 in W. Brian Arthur, Steven N. Durlauf, and David A. Lane, The Economy as an Evolving Complex System II, Santa Fe Institute Studies in the Sciences of Complexity, Vol. XXVII, Addison-Wesley, 1997. [29 páginas]
brian arthur price assets ahlpt96.pdf This paper develops the Santa Fe Artificial Stock Market Model. 7.10. "Complexity and the Economy," Science, 2 April 1999, 284, 107- 109. [5 páginas]
bbrian arthur Econ_&_Complex_Web.pdf 7.11. "The End of Certainty in Economics," Talk delivered at the conference Einstein Meets Magritte, Free University of Brussels, 1994. Appeared in Einstein Meets Magritte, D. Aerts, J. Broekaert, E. Mathijs, eds. 1999, Kluwer Academic Publishers, Holland. Reprinted in The Biology of Business, J.H. Clippinger, ed., 1999, Jossey-Bass Publishers.[6 páginas] vid p. 87 supra. 7.12. "Cognition: The Black Box of Economics," The Complexity Vision and the Teaching of Economics, David Colander, ed., Edward Elgar Publishing, Northampton, Mass, 2000.[7 páginas]
brian arthur Colander_Cognition_Web.pdf 7.13. "Myths and Realities of the High-Tech Economy," Talk given at Credit Suisse First Boston Thought Leader Forum, Sep 10, 2000. [5 páginas]
brian arthur Credit_Suisse_Web.pdf 7.14. “Is the Information Revolution Dead? If history is a guide, it is not.”, Business 2.0, March 2002 Issue. [8 páginas]: vid p. 93 supra
102 103
W. Brian Arthur
Some Selected Papers http://www.santafe.edu/arthur/Papers/Papers.html
"Complexity and the Economy," Science, 2 April 1999, 284, 107-109. PDF MS Word
"The End of Certainty in Economics," Talk delivered at the conference Einstein Meets Magritte, Free University of Brussels, 1994. Appeared in Einstein Meets Magritte, D. Aerts, J. Broekaert, E. Mathijs, eds. 1999, Kluwer Academic Publishers, Holland. Reprinted in The Biology of Business, J.H. Clippinger, ed., 1999, Jossey- Bass Publishers. HTML PDF MS Word
"Positive Feedbacks in the Economy," Scientific American, Feb. 1990. Pdf
"Competing Technologies, Increasing Returns and Lock-in by Historical Events," Economic Journal, 99, 106-131,1989. Pdf
"Increasing Returns and the New World of Business,"Harvard Business Review, July-Aug 1996. Pdf MS Word
"Bounded Rationality and Inductive Behavior (the El Farol Problem), American Economic Review, 84,406-411, 1994. Html Pdf
"Complexity in Economic and Financial Markets," Complexity, 1, 20-25, 1995. Pdf
Preface to the book: Increasing Returns and Path Dependence in the Economy, Univ. of Michigan Press, Ann Arbor, 1994. Pdf
"Process and Emergence in the Economy," introduction to the book The Economy as an Evolving Complex System II, edited by Arthur, Durlauf, and Lane, Addison Wesley, Reading, Mass, 1997. Html Pdf
"Cognition: The Black Box of Economics," The Complexity Vision and the Teaching of Economics, David Colander, ed., Edward Elgar Publishing, Northampton, Mass, 2000. PDF MS Word
"Myths and Realities of the High-Tech Economy," Talk given at Credit Suisse First Boston Thought Leader Forum, Sep 10, 2000. PDF MS Word
"The Economy as an Evolving Complex System II.," W. Brian Arthur, Steven N. Durlauf, and David A. Lane, (Eds.), Proceedings Volume XXVII, Santa Fe Institute Studies in the Science of Complexity, Reading, MA: Addison-Wesley, 1997. Review by Gerald Silverberg, Maastricht. PDF MS Word
Last Modified: Monday, December 17, 2001
103 104
"Is the Information Revolution Dead?" Business 2.0, March 2002. MS Word
"Complexity and the Economy," Science, 2 April 1999, 284, 107-109. PDF MS Word
"Cognition: The Black Box of Economics," The Complexity Vision and the Teaching of Economics, David Colander, ed., Edward Elgar Publishing, Northampton, Mass, 2000. PDF MS Word
"Myths and Realities of the High-Tech Economy," Talk given at Credit Suisse First Boston Thought Leader Forum, Sep 10, 2000. PDF MS Word
"The End of Certainty in Economics," Talk delivered at the conference Einstein Meets Magritte, Free University of Brussels, 1994. Appeared in Einstein Meets Magritte, D. Aerts, J. Broekaert, E. Mathijs, eds. 1999, Kluwer Academic Publishers, Holland. HTML PDF MS Word
Last Modified: Thursday, April 25, 2002
104 105
Biographical Information
A Short Bio-Background
Business Résumé PDF MSWord
A full publication list is available on request.
Additional bio material available in Who's Who in America, Who's Who in Economics
For a history of my early work in increasing returns and complexity and role in getting Santa Fe Institute's first research program started, see M. Mitchell Waldrop's book Complexity, Simon and Schuster, 1993; and the interesting article on publication practices in economics: How are the Mighty Fallen: Rejected Classic articles by Leading Economists, by Gans and Shepard, J. Econ Perspectives 8, 165-179, 1994
For the connection between early increasing returns work of mine and the current Dept. of Justice vs. Microsoft case, see John Cassidy's article, "The Force of an Idea," in the New Yorker, January 12, 1998.
Last Modified: Monday, December 17, 2001
Web Links
The Santa Fe Institute
105 106
Selected Publications
Books
The Economy as an Evolving Complex System II. Edited (with S. Durlauf and D. Lane), Addison-Wesley, 1997. Increasing Returns and Path-Dependence in the Economy, University of Michigan Press, Ann Arbor, Mich., 1994. The Economic Consequences of Changing Age Distributions. Edited (with R.D. Lee and G. Rodgers), Oxford University Press, 1987. Population, Development and Food. Edited (with R.D. Lee, T.N. Srinavasan, and G. Rodgers), Oxford University Press, 1988.
Selected Papers and Articles “Cognition: the Black Box of Economics,” The Complexity Vision and the Teaching of Economics, D. Colander, ed., 2000, Edward Elgar Publishers. “Time Series Properties of an Artificial Stock Market,” with B. LeBaron and R. Palmer, Journal of Economic Dynamics and Control, 23, 1487-1516, 1999. “Complexity and the Economy,” Science, 2 April 1999, 284, 107-109. Reprinted in The Complexity Vision and the Teaching of Economics, D. Colander, ed., Edward Elgar Publishers, 2000. “Asset Pricing Under Endogenous Expectations in an Artificial Stock Market,” with J.H. Holland, B. LeBaron, R. Palmer, and P. Tayler, SFI Paper 96-12-093, Economic Notes. Reprinted in The Economy as an Evolving Complex System II. Edited (with S. Durlauf and D. Lane), Addison-Wesley, 1997. “Beyond Rational Expectations: Indeterminacy in Economic and Financial Markets” in Frontiers of the New Institutional Economics, J.N. Drobak and J.V. Nye (eds.), Academic Press, 1997. “The End of Certainty in Economics,” in Einstein meets Magritte, D. Aerts, J. Broekaert and E. Mathijs, eds. 1999, Kluwer academic Publishers, Holland. Reprinted in The Biology of Business, J.H. Clippinger, 1999, Jossey-Bass Publishers. “How Fast is Technology Evolving?” Essay in Scientific American, Feb 1997. “Increasing Returns and the New World of Business,” Harvard Business Review, July-Aug 1996. Reprinted in: The Diamond Weekly, Japan; in The Strategist Quarterly, India; in Seeing Differently, John Seely Brown, Harvard Business School Press, 1997; in The Knowledge Economy, Dale Neef, ed. 1997, Butterworth-Heinemann, Boston. “Complexity in Economic and Financial Markets,” Complexity, 1, 20-25, 1995 “Artificial Economic Life: a Simple Model of a Stockmarket,” with R. Palmer, J. Holland, B. LeBaron, and P. Taylor, Physica D, 75, 264-274, 1994 “Inductive Reasoning and Bounded Rationality,” American Economic Review, (A.E.A. Papers and Proc.), 84, 406-411, 1994 “On the Evolution of Complexity,” in Complexity, G. Cowan, D. Pines, D. Melzer (eds.) Addison-Wesley, 1994
106 107
“Economic Agents that Behave like Human Agents,” Journal of Evolutionary Economics, 3, 1-22, 1993 Reprinted in The Legacy of Joseph A. Schumpeter, H. Hanusch, Ed., Edward Elgar Publishers, 2000 “Information Contagion,” with David A. Lane, Structural Change and Economic Dynamics, 4, 81-104, 1993 “Why do Things become more Complex?” Essay in Scientific American, May 1993 “Dynamic Equilibria in Markets with a Conformity Effect,” with Andrzej Ruszczynski, Archives of Control Sciences, 37, 7-31, 1992 “Learning and Adaptation in the Economy,” Santa Fe Institute Paper 92-07-038 , 1992 “Dynamics and Structures,” with Michael Landesmann and Roberto Scazzieri, Structural Change and Economic Dynamics, 2, 1-7, 1991. “On Classifier Systems and Models of Learning in the Economy,” mimeo, Santa Fe Inst, 1990. “Designing Economic Agents that Act like Human Agents: A Behavioral Approach to Bounded Rationality,” American Economic Review (A.E.A. Papers and Proc.) 81, 353-359, 1991. Reprinted in The Economic Legacy of Robert Lucas, Jr. K.D. Hoover, Ed. 1998, Edward Elgar Publishers "Positive Feedbacks in the Economy," Scientific American, 262, 92-99, Feb. 1990 “A Learning Algorithm that Mimics Human Learning,” Santa Fe Institute Paper 90-026, 1990 “Silicon Valley Locational Clusters: When do Increasing Returns Imply Monopoly?" Mathematical Social Sciences, 19, 235-251, 1990 "The Economy as a Complex System," in Complex Systems, D. Stein (ed.), Wiley, New York., 1989 "Nash-Discovering Automata for Finite-Action Games," Mimeo. Santa Fe Institute, 1989 "Self-Reinforcing Mechanisms in Economics," in The Economy as an Evolving Complex System, pp 9-33, K. J. Arrow and P. Anderson (eds.), Wiley, New York., 1988 "Competing Technologies, Increasing Returns, and Lock-In by Historical Events," Economic Journal, 99, 116-131, 1989, (IIASA Paper WP-83-90, September 1983) Reprinted in: The Economics of Technical Change and International Trade, G. Dose, K. Pavitt and L. Soete, eds. 1991, New York Univ. Press, N. York; in Technical Change and Economic Theory, G. Dosi, ed, 1992, IFIAS Research Series; in The Economics of Productivity, E. N. Wolff, 1997, Edward Elgar Publishers; in Economics of Innovation, Christopher Freeman, ed. 1997, Edward Elgar Publishers; in Market Process Theories, P.J Boettke and D. L. Prychitko, eds. 1998, Edward Elgar Publishers; in The Economics of Increasing Returns, ed. G. Heal, 1999, Edward Elgar Publishers. "Immigration Policy and Immigrants' Ages," (with T. J. Espenshade), Population and Development Review, 14, 315-326, 1988 "Adaptive Growth Processes Modeled By Urn Schemes," (with Yu. M. Ermoliev and Yu. M. Kaniovski) Kibernetika , No.6, 49-57, 1987. Translated in Cybernetics, 6, 1987 "Urban Systems and Historical Path-Dependence," Chapt. 4 in Urban systems and Infrastructure, J. Ausubel and R. Herman (eds.), National Academy of Sciences, Washington, D.C., 1988 "Competing Technologies: An Overview," Chapt. 26 in Technology and Economics, G. Dosi, C. Freeman, R. Nelson, L. Soete, G. Silverberg, (eds.), 1988 "Non-Linear Urn Processes: Asymptotic Behavior and Applications" with Yu. M. Ermoliev and Yu. M. Kaniovski, I.I.A.S.A. Working Paper 87-85, 1987 "Strong Laws for a Class of Path-Dependent Urn Processes," (with Yu. M. Ermoliev and Yu. M. Kaniovski ) in Proc. International Conf. on Stochastic Optimization, Kiev 1984, Springer, Series Info. and Control, 81, 1986
107 108
"Path-Dependent Processes and the Emergence of Macro-Structure" (with Yu. M. Ermoliev and Yu. M. Kaniovski ), European Journal of Operations Research, 30, 294-303, 1987, "Some General Relationships in Population Dynamics" (with James W. Vaupel), Population Index, 50, 214- 255, 1984 "Competing Technologies and Economic Prediction," Options, April 1984, 10-13 Reprinted in The Social Shaping of Technology, D. MacKenzie and J. Wajcman, Open University Press, 1999. "A Generalized Urn Problem and its Applications" (with Yu. M. Ermoliev and Yu. M. Kaniovski) in Kibernetika , 19, 49-56, 1983. Trans. in Cybernetics, 19, 61-71, 1983 "An Analysis of Indirect Mortality Estimation" (with Michael Stoto), Population Studies, 37, 301-314, 1983 "Age and Earnings in the Labor Market: Implications of the 1980's Labor Bulge," Chapter 21 in Human Resources, Employment, and Development, Vol. 2, Paul Streeten and Harry Maier, eds., Macmillan, 1983 "The Analysis of Linkages in Demographic Theory," Demography, 21, 109-128, 1984 "The Ergodic Theorems of Demography: A Simple Proof," Demography, 19, 439-445, 1982 "Immigration and the Stable Population Model," (with Thomas Espenshade and Leon Bouvier), Demography, 19, 125-133, Feb. 1982 “The Economics of Risks to Life,” American Economic Review, 71, 54-64, 1981 "Why a Population Converges to Stability," American Mathematical Monthly, 88, 557-563, 1981 "An Analytical Survey of Population and Development in Bangladesh" (with Geoffrey McNicoll), Population and Development Review, 4, 23-80, 1978 "Stochastic Control of Linear, Discrete-Time, Distributed-Lag Models," International Journal of Control, 28, 611-619, 1978 "Samuelson, Population and Intergenerational Transfers" (with Geoffrey McNicoll), International Economic Review, 19, 241-246, 1978 "Control of Linear Processes with Distributed Lags Using Dynamic Programming from First Principles," Journal of Optimization Theory and Applications, 23, 429-443, 1977 "Optimal Time Paths with Age-Dependence: A Theory of Population Policy" (with Geoffrey McNicoll), Review of Economic Studies, XLIV (i), 111-123, 1977 "Large-Scale Simulation Models in Population and Development: What Use to Planners?" (with Geoffrey McNicoll), Population and Development Review, 1, 251-265, 1975 "Optimal Control Theory with Time Delay" (Ph.D. Thesis), Operations Research Center Report , 73-37, University of California, Berkeley, 1973 December 2000
108