What's the Right Economics for Cyberspace? by Michael H. Goldhaber
First Monday
What's the Right Economics for Cyberspace?  by Michael H. Goldhaber

In April's First Monday , I argued that the growth of cyberspace heralds  a new kind  of economy, and with it , a new kind of economics, both based on the characteristics not of material, mass-produced goods, nor of money, nor of information, but of attention. In a reply, one of the editors of this journal, Rishab Aiyer Ghosh, argued against the  very possibility that  the "law of supply and demand" and other basic laws of classical economics might no longer hold.  He took me to task for making this claim, and tried to show it was wrong in the case of attention.

In this rejoinder to Ghosh, I show that the assumptions of classical and neo classical economics are inappropriate for the attention economics that the  orginal article introduced. Attention cannot  meaningfully be quantified, even approximately, thought the total available is still limited.  Ghosh's main points are shown to depend on ignoring this key fact. I also offer further evidence and argument that  standard economics fails for the net.



A Different State of Affairs
What Rishab Ghosh Thinks
Why Attention is Not a Commodity
On the Margins with Oprah and Me
Sad Times at the (Attention) Supermarket
Die-Hard Fans
How the Old Economics Has Already Failed on the Net

Which economic theory best describes what goes on on the Internet? If you believe as I do that the explosive growth of the net and its relatives will likely continue until they become the dominant arenas for human effort and involvement, this is no idle question. Theories are guides to action: sticking to the wrong one can lead to ruin; adopting the right one opens the path to success.

 Given the importance of the question, then, it is a little surprising that most people simply take for granted that the correct answer is neo-classical economics. They assume without any basis that this approach has to work as well on the net as it is currently thought to do in other parts of life. They ignore that every known theory, of whatever nature, from physics to sociology, is developed and tested within a certain domain of facts, and cannot be assumed automatically to be valid in any different domain. In some cases a theory does turn out to work when extended; in other cases, not at all. Without some kind of probing and testing you just can't know, not even when your choice has been extremely successful so far. There is no reason whatsoever to take economic theory to be an exception in this.

 Suppose archaeologists were to unearth the remains of a previously unknown civilization; should they assume from the outset that neo-classical economics correctly explained what went on there? Obviously not. Much the same applies to cyberspace, including the Net. In effect it is a newly discovered civilization, despite the fact that we ourselves and our descendants will inhabit it. Without examination, nothing justifies concluding in advance that neo-classical economics will apply. Indeed, because life in cyberspace is so different from what we are used to outside it, as a first guess, it makes sense to tilt against that supposition.

 These thoughts were part of what motivated the investigations I summarized in my recent article "The Attention Economy." There I argued two main points. First, the economy that works on the Net revolves around the scarcity of attention. Second, this economy operates differently from any predicted or accommodated by neo-classical economic theory. If need be, readers should return to that article to acquaint themselves with the reasoning behind my first point. As for the second one, part of my argument there boils down to the following. As I demonstrate, the natural concept of property on the Net is attention, which is held in the minds of one's beholders. This is utterly different from intellectual property, which is the form that follows very directly from the neo-classical theory. If two theories lead to different conclusions, they must be different theories. Thus, if you accept the importance of attention, you are forced to agree that cyberspace entails a different kind of economics.



A Different State of Affairs

Attention economics on its own is nothing new, even though perhaps I have noticed it more explicitly than anyone before me. For thousands of years, as long as there have been conversations or large audiences focusing on just a few people, this peculiar economics has been in play. However, in the past, it has rarely been dominant. Further, standard economists have never developed much of a theory of it, despite at least a century available to them to do so. I suspect the reason they didn't was that they would have had to abandon the classical, Keynesian, neo-classical or even Marxist versions of economics they developed in the industrial era.

 Attention is nothing simple. It is intricate, individual, variegated, quite often unpredictable - anything but uniform, measurable and neatly quantifiable. It entails all the senses; it shades off into recognition, memory, appreciation, reflection, love, understanding, empathy, observation, care, attentiveness, waiting, serving and more - all of which are messy and hard to quantify. The attention I am capable of paying is different from yours; the kinds of attention I desire differs from what you want as well. In neither case is it likely that there can be a ranking between us - you more attentive in general, me less, nor vice versa. Nor are we completely equal.

 So for the most part attention resists the kind of mathematical treatment that went so well with the consistent, uniform, interchangeable, generally anonymous and impersonal products, jobs and money which formed the core of the industrial era. Still the sum total of attention available to everyone comes from the sum total of attention everyone can pay, and that is limited. It is precisely because the attention concept is so richly complex yet the available totality of attention is limited that attention can serve as the full basis of an economy. But it is an economy unlike what we are used to; it has to be.

 Attention is like a resource only in the most metaphoric sense, though "resource," for lack of a better term is useful in thinking about it; the same goes for "scarcity." Attention is not standardized and measurable like a commodity; you can't give or trade away all the attention you have accumulated, as would be possible with currency and the goods bought with it, nor can you barter attention for something else. Nonetheless, when you have sufficient attention (say as a star) you can get people to pay you still more: if you want they will do this by doing just about anything for you or supplying you with anything they can make or get their hands on.

 If attention is viewed as a resource and I "consume" some, then my audience is its source. They are "producing" it, in standard economic terms. However, economists usually speak of membership in an audience as a form of consumption, not production. Furthermore, audiences do turn out to be one of the few important kinds of entity that are now truly scarce, but yet, today, very commonly members of audiences are still called on to pay for the privilege of paying attention. Altogether then, distinctions vital to classical and neo-classical economics no longer turn out to be very helpful. A different set of concepts is clearly called for.


What Rishab Ghosh Thinks

I welcome Rishab Aiyer Ghosh's response to my article, not only because he does accept the validity of my first point, the importance of attention in understanding the economy of cyberspace, but just as much because the rest of his reaction provides almost a textbook example of how not to think about the economics of attention. Ghosh's primary allegiance appears to be to what he terms "classical economics." [ 1 ]  Its truth, he implies, is outside history altogether. We have always lived, it would seem, in what he calls "the economy" and apparently always will. At most, economic theory can undergo only minor changes. He compares it with geometry in the absoluteness of its truth.

 It is fun then, to consider Ghosh's remarks about geometry. He correctly points out that when you depart from the familiar Euclidean geometry many of us studied in high school, to the geometry of a sphere (which he incorrectly labels "Riemannian" though it was worked out well before Riemann) you find that the angles of a triangle no longer have to add up to a half circle, but it remains true that the sum of the lengths of any two sides will be greater than the length of the third side. Had he gone on to consider the well-known Riemannian geometry of four-dimensional space time, in which, according to Einstein we actually live, he would have found that it is easy to construct a triangle where the length relation he touts no longer holds. You could go further, into topology, in which length itself loses all meaning, and still have a valid theory of space for certain classes of problem.

 There is an interesting analogy; as we experience cyberspace as space, it is very different from the space of the industrial market system. In the latter, distances are measured in miles or kilometers, and you have no direct access to what goes on far away. In cyberspace, attention flows from site to site, and the distance between sites can only be measured in terms of how many intermediate sites must be visited to pass from one to the other. That in turn depends on how much attention those who maintain each site offer to those who maintain the other sites. The more they offer, the more likely they are to allow a visitor to click on the relevant URL, so the less the separation. Thus the geometry of cyberspace is nothing like the geometry of the prior economy, which offers at least a hint about the respective economic laws as well.


Why Attention is Not a Commodity

In trying to hold to his faith that standard economic theory must hold, Ghosh makes a number of arguments in which he repeats the same sort of mistake over and over. The mistake is that attention must be commodity-like. Used by economists the word "commodity" most commonly refers to something uniform and interchangeable, so that it may be purchased easily at a standard price without regard for its origins. Examples include durum wheat, aviation fuel, 20-gauge rolled steel, AA batteries, folding chairs, microwave ovens, long-stemmed roses, or 1-Megabyte EDO RAM chips. More loosely, of course, "commodity" just means something that can be sold, or more loosely still, it just means something that is somehow valued, like love. Only in the last sense is the term applicable to attention, and that usage is far to loose to afford any theoretical application. Ghosh simply glides over these important distinctions of meaning.

 If attention were a commodity in a strict sense it could be bought. Consider the case where it is supposedly offered by, as Ghosh would have it "those who politely point you to the early music section in a store." Suppose you walk into the store, having an early jazz performance in mind. Without fully understanding you or hearing you out, the clerk "politely points you to the early music section," which is completely the wrong place.

 This is not attention but inattention, though superficially it looks the same. The mere fact that you may be a customer, or even have been one before is no guarantee that you will get the attention you want under the circumstances. If the service in question is actually for sale, as is the case, say, with a massage or a taxi ride, things are no better; you still cannot be sure how you will be treated or listened to based on what you pay. On the other hand, if you are sufficiently well-known, interesting or attentive to the service provider, your chances greatly improve, regardless of whether you pay money or not.

 Suppose you go further, taking the ultimate step of paying an entire audience to be attentive to you. In that case, you still have to keep them interested once you have paid. Otherwise they will sit there musing about something else, or perhaps fall asleep. In other words, attention just cannot be reliably bought for any specific price. It is not a commodity.


On the Margins with Oprah and Me

In another passage Ghosh urges that there must be some value in comparing the "marginal utilities" that I and the super well-known talk show host Oprah Winfrey would discern in adding certain numbers of members to our respective audiences. Ghosh forgets that marginal utility is a concept abstracted from a world in which standardized goods are bought and sold for money or something very much like it. Two people's marginal utility can be compared only if it is measured in the same way, usually, in practice by the amount of money they are willing to pay for something whose utility is at issue. If there is no common currency between us, it is impossible to compare my subjective feelings about attention with Oprah's. But if you cannot buy attention, it is quite unclear what that common currency is to be.

 This is not to say that Oprah and I could not, conceivably enter into an attention transaction. It is improbable, but possible that she would have me on her show, perhaps to tout my book on this subject, once it's published. In that unlikely event, I would gain some small amount of attention from her large audience; it would then be up to me to hold that attention. Presumably, all my fans (all three of them, that is) would tune into Oprah to see me, (for their first time, I would guess) and, not inconceivably, they would like what they saw enough to turn into Oprah fans as well.

 Now, of course, Oprah is unlikely to invite me because she would have little reason to believe that most of her fans would appreciate hearing what I have to say, and the paltry few additional fans that might come to her by doing this are not worth even slightly alienating the millions she has. Undoubtedly, you could if you wanted, construct this argument in terms of marginal utilities, but you would not thereby gain the theoretical force the concept has in industrial economies. Instead, it would just be an exercise leading to a conclusion that was obvious to begin with.


Sad Times at the (Attention) Supermarket

In a third passage, Ghosh wonders whether attention can be a currency. He seems to believe that I argued it could be traded, and that that means I accept that there could be a market for attention. This involves a misunderstanding of how standard markets work. Markets usually allow prices to reach some equilibrium, because buyers and sellers can compare different offers. When I exchange attention for attention, what different offers can I be comparing? The attention of a friend tends to be more valuable to me than that of a stranger, because I have past experience with this same friend. We have built up our memories of one another through paying attention over time. If some stranger offered to pay me twice as much attention as my friend would for the same amount of mine, I could not evaluate this offer. Only by actually going through with the experience of paying the attention to the stranger and so discovering what I felt I was getting in return, could I possibly decide that her attention was more worthwhile than my friend's. There can be no bid and counterbid in advance of a transaction, and thus nothing resembling the equilibrium that characterizes standard markets in action. Thus, in a conventional sense there is no market.

 Attention, as I mentioned before, cannot be a currency, because for that I would have to be able to spend all the attention I had managed to accumulate, in trade for something else. But that is impossible. The attention I have resides in others' minds, and I cannot perform an operation to make them believe, say, that all the time they were paying attention not to me but to you. A currency that cannot be handed over at will cannot be spent at will and so is not much of a currency, certainly.


Die-Hard Fans

In yet another passage, Ghosh argues:
if you watch Die Hard I, II, III, IV and V, you'll be less bored with Bruce Willis than if you watch Die Hard 1 to 365 every day. Quite possibly, for the average viewer, Die Hard V watched the tenth time will be less interesting than when watched the ninth time. Samuelson again: "the law of diminishing marginal utility states that, as the amount of a good consumed increases, the marginal utility of that good tends to diminish."

 Goldhaber agrees: "since it is hard to get new attention by repeating exactly what you or someone else has done before, this new economy is based on endless originality." But instead of mentioning this as an instance of supply, demand and marginal utility operating on the scarce resource of attention, leave alone further exploring the topic with this tool of classical economics, he contrasts it with the mass production of industrial economies."

 But, to begin with, Ghosh completely misunderstands what it would mean if you watched Die Hard 1 to 365 every day (assuming they existed and were possible to watch in so short a time). You would be intensely inclined to watch Die Hard 366, for one thing. You would also be more, not less inclined to want to know everything you could about Bruce Willis. You would undoubtedly favor this actor and this series over anything that seemed superficially comparable.

 However, the main reason that such a long series is unlikely is that the verve and vitality of the first in a series is hard to sustain, on the part of its creators. Think in terms of illusory attention, the feeling in an audience member that the stars - here the movie makers as well as the chief actors - are paying attention to her. For this they have to evince by thier actions that that they really mean it, and it is hard to mean exactly what you said before. Samuelson's law of diminishing returns, which holds in all situations - except when it doesn't - offers no great insight here. Believability is much more the issue.

 The impetus for the making of a series of movies following on an original success is the purely commercial one of perfecting routine, building a business by making more of the same, which is intrinsic to the old economy. The qualitative kinds of distinction attention economics would make here cannot be captured in Samuelson's quantitative formulae, and if Ghosh were not such a fan of the latter perhaps he would see that more readily.

 Another actor who is "just like" Willis on the screen (does that mean looks the same, makes each gesture the same, says his lines exactly the same?) is impossible to envisage. You watch a star not only for what she is doing at the moment, but for her entire biography. Willis acted opposite Cybille Shepard in a long-running detective sitcom; he is married to Demi Moore; he appeared in Quentin Tarantino's Pulp Fiction, to take some facts that I - no great fan - happen to be aware of. While I have little interest in the sort of movie I imagine Die Hard (I) to be, the fact Willis starred in it at least made me think about going to see it. Suppose someone made a movie that could be described in very similar terms, yet without a star that fans had already paid attention to in the past or found utterly entrancing in watching brief previews. It would be unlikely to attract much of an audience. Theories of supply and demand are a hindrance in grasping this, not a help, as far as I can make out. Again, Samuelson 's quantitative relationships have nothing to offer about the singularity of a person, or what distinguishes one actor from another, and that is precisely what is at stake here.


How the Old Economics Has Already Failed on the Net

Ghosh is one of the editors of First Monday. For that reason he knows very well that in practice classical economics has so far failed in regard to the Net. When the Web first appeared on the scene, it was widely heralded as a major future source of revenues, especially for some of the key "industries" of this post-industrial period, from software to book, periodical, and music publishing, to movie-making and broadcasting, among others. It was evident at once that in future their "output" was going to have to be made available over the Web. But what turns out to be more difficult to envision is how they are going to profit from this monetarily - or just come out even.

 Almost everyone, including the editors and publishers of this very journal, seem to have come to the conclusion by now that the way to make money on the Net is not by selling a chance to pay attention to what one produces, as they originally thought in effect, but by selling advertisers the chance of getting attention though close association on the computer screen [ 1 ]. In other words, they now view the Internet as merely a new mass medium, and assume it to be open to the same kinds of tactics that seemed to work well for print and broadcast.

 For these tactics to work on the Net, however, people must spend money in sufficient amounts for what is advertised. Consider what would be at stake for First Monday. The kinds of advertisers that in the past supplemented the revenues of comparable printed journals of ideas and scholarship have mostly been publishers of books. It won't be long, presumably, before books too will appear in web versions, so that they too can no longer be sold in the long-accepted way, and thus would also have to depend on advertising.

 Ultimately, all the Web and Net advertising will have to be supported by payments made for the kinds of things that cannot be obtained directly through cyberspace. That means material objects and services such as computers, breadsticks, toothpaste, or airline trips. If the Net and other attention-related activities continue to grow in importance (which I would strongly argue the available evidence points towards) then the sum total of material products and services together will eventually account for only a small fraction of human effort. By then, as a result, the vast majority of the monetary cost of buying a Web-advertised product would simply go to pay for the advertising. (Otherwise most of what occurs on the Web would already be outside the money economy.) Why would anyone be moved to pay such premium prices as that suggests? Similar but little-advertised versions of the same things would be much, much cheaper, while knowledge of these lower-priced alternatives could also, ironically, be easily obtained on the Net itself.

 Two possible conclusion may be drawn, still basically within the framework of standard economics. The growth of the Web, including this journal and most other sites will not be sustainable by advertising, or else the whole system will operate in an increasingly irrational manner. Web users would knowingly pay gigantic premiums for advertised products in order to sustain the Web, or perhaps from being incredibly seduced by advertising. However, the latter state of affairs would better be described as an elaborate charade. We would all be conspiring together to deny the real workings of the new system, where prices, money and profits in fact pay little role, in which there is no real market, in which the economic issues that dominate thinking now would have no meaning or relevance. These are issues such as GNP, unemployment rates, the swings of capital markets, interest rates, money supply, exchange rates, taxes and a vast array of other concepts developed by neo-classical economics over the years - yes, concepts including the law of supply and demand and concepts such as marginal utility.

 If I am remotely right, we have entered into a transition to a vastly different system which will come to dominate all aspects of life, probably irrevocably. There is much more to do in working out how this system will function. If Ghosh (or any reader) is interested in trying to discover ways in which aspects of classical or neoclassical economics might be salvaged for use in the attention economy, he is more than welcome. But there is no reason for him to scold me for not being entranced by this issue. Still less is he justified in criticizing my meager efforts in alerting the world that, contrary to the overwhelming received opinion, standard economics cannot be relied on in this new realm.End of article

The Author

Michael H. Goldhaber is completing a book on the attention economy. Formerly a theoretical physicist, a Fellow of the Institute for Policy Studies in Washington, D. C. and editor of Post-Industrial Issues, he is currently head of his own think tank, The Center for Technology and Democracy, and is a visiting scholar at UC Berkeley's institute for the Study of Social Change. is previous book was Reinventing Technology.

His Web site is http://www.well.com/user/mgoldh/

 E-mail: mgoldh@well.com
©Michael H. Goldhaber, 1997



1. Ghosh's problem with what I write originates in his disagreement over my assertion that the law of supply and demand is outmoded in the new economy. This law is usually taken to be a general rule about  prices: For each kind of good prices rise when demand rises or supplies fall; they fall when supply rises or demand falls. If there is nothing equivalent to price, the statement has no obvious import. Ghosh never states what he thinks the law is, but he apparently believes it to be something different and more general, so much so that one cannot talk about the importance of scarcity without implying it. At this level, his argument with me over my reference to the inapplicability of the law boils down to nothing but semantics.

When I first hit upon the imporatance of attention I was indeed guided by the broad idea that economics relates to scarcity. Later, as I proceeded to investigate just how attention works, I was forced to ask myself in more general terms why scarcity mattered. The answer has to do not with any quantitative formula, but rather with the straightforward thought that if what you desire is scarce and you think you stnd a chance of getting it, you will  quite possibly engage in efforts to obtain it. Your effort will influence and be connected with other's efforts, and connected effort is what an economy, at bottom, is about. This formulation does not appear to me to be  closely  related to the law of supply and demand as it is commonly understood. It has nothing to do with the behavior of prices. It doesn't suggest that if you want attention more than I do that you will devote more effort to getting it. Thus, as I understand the concpet of a "law," namely some formula that always applies or applies on the average, there is no law of supply and demand operating,nor anything close to one. However, nothing prevents Ghosh or anyone else from defining this law differently from what is commonly understood, and then insisting her definition is right.

2. Parethetically, Ghosh believes that advertising offers an instance of attention being "monetized." If I understand what he means by this, it's wrong. Advertisers pay, at best, for a crack at attention, something altogether different from attention itself.
They do this, for instance, when they buy expensive commercial time on a successful TV program. Their hopes are that by making the ad itself attention-getting enough they can keep the audience from pressing the mute button, going to the bathroom or refrigerator, thinking about something else, or flicking to another channel.

Copyright © 1997, First Monday

What's the Right Economics for Cyberspace? by Michael H. Goldhaber.
First Monday, Volume 2, Number 7 - 7 July 1997

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2020. ISSN 1396-0466.