DRAFT: From þe "Commodity Economy" to þe "Attention Economy"?
A piece I need to develop much more; and for which I will have a very hard time indeed finding an outlet...
The 2020-2022 COVID-19 plague caused by the coronavirus SARS-COVID-19 is, many today think, going to place the world economy in an odd position. For more than a quarter century now the ongoing revolution in data processing and data communications technology has been profoundly altering the basic underpinnings of property and exchange and value—has been shifting us from a “commodity” to an “attention” economy. We have all watched this shift gradually lurch forward. But now, many say, our two plague years are packing, in some aspects, what would have been at least a decade of change and development into two years, as requirements of quarantine and social distancing have taught and humans have had to learn on an emergency basis how to better use information and communications technology to try to substitute for much face-to-face interaction.
I am, today, here to think out loud about whether this is true. If it is true, what it means. If it means much of anything, how should we react.
Let me first attempt to set out two ideal types: that of a commodity economy—as set out by Adam Smith two and a half centuries ago in his Wealth of Nations—and that of the attention economy toward which we have been moving.
In a commodity economy, in my image of it at least, things of value are concrete easily assayable commodities created for and then distributed via sale in competitive markets. The commodities are rival, excludible, transparent, and as for coordination, market prices creating incentives directing small-scale producers are enough to solve coordination problems. In fact, the properly tuned market economy solves all the problems:
Because in a commodity economy the principal sources of value are rival goods, requiring people to pay prices for using them is a very good way to incentive-compatibly crowdsource the solution to the problem of how what is produced is going to be distributed:
[willingness to pay] = [intensity of demand] x [social power (mediated through their wealth) of the demander].
That use of goods is also easily excludible makes it possible to implement such a price and market exchange-based system. Then, in addition, it immediately drops out that you have solve the problem of what should be produced: whatever makes a profit should be produced. And as long as you are comfortable with the distribution among people of social power implicit in the distribution of wealth, and as long as Pigouvian externalities are not widespread—and you can at least argue in favor of the powerful insights of Ronald Coase (or is it a dodge?) that it is possible to “cut property rights at the joints”, so to speak, to all but eliminate transaction costs and so make externalities truly de minimis—property, contract, and the market are what you need. What then is not to like?
Well, there may be information problems—adverse selection and moral hazard. Well, there is the question of whether the market and its prices are a thick enough communication channel to actually do the coordination of human activity required to make production efficient. Well, there is the question of who is going to do the R&D, and what is the right amount of R&D to do. Well, there is the question of market power.
Of these the last, the problem of market power, was well-known to Adam Smith’s generation. It was a major wellspring of his belief that only landlords should have the vote. The working class—even if a well-functioning commercial society made it prosperous and comfortable, and even if the government took a hand to moderate the stultifying effects of the division of labor on the minds of the working class, and make no mistake: Smith thought a sophisticated division of labor was both marvelously productive and appallingly stultifying—could not handle the required cognitive load of voting. And the mercantile and manufacturing classes would, if given the vote, use it to have the government give them all monopolies. But landlords, Smith thought, had an interest both in general prosperity and in competitive markets.
Of these the next to last, the problem of R&D, was always an embarrassment.
At the start it was a minor embarrassment. Suppose that we are willing to make the brave guess that the efficiency-of-labor E in a Solow growth model is equal to the value of the stock of useful ideas discovered, developed, and deployed in the world economy H times the square-root of natural resources R per worker L:
Then in Adam Smith’s age, from 1500 to 1770, worldwide, the proportional rate of growth h of the value of the deployed-ideas stock H was 0.15%/year: less than 5% per generation; only 16% per century. Adam Smith saw a world in which the density and sophistication of your commercial network and thus the fineness of your societal division of labor was much more important as a determinant of the efficiency-of-labor than was the process of invention, innovation, development, and deployment. Indeed, he thought of those largely as immediate consequences of a fine division of labor, rather than as an independent force.
It was soon thereafter clear that that judgment was wrong for the future humanity was then moving into. Over 1770-1870 our h becomes 0.45%/year, worldwide. And starting in 1870 it has been approximately 2%: 4.5 times its pace during the 1770-1870 Industrial Revolution century, 14 times its pace during the 1500-1770 Imperial-Commercial Revolution era, (and 55 times its pace during the preceding 1-1500 Late Agrarian Age). By the 1840s Friedrich Engels was dismissing classical economists as obvious paid shills for the rich because of their reduction of:
the production costs of a commodity... [to] three elements: the rent... the capital with its profit, and the wages…. A factor which the economist does not think about–I mean the mental element of invention, of thought…. What has the economist to do with inventiveness? Have not all inventions fallen into his lap without any effort on his part? Has one of them cost him anything? Why then should he bother about them in the calculation of production costs? Land, capital and labour are for him the conditions of wealth, and he requires nothing else. Science is no concern of his. What does it matter to him that he has received its gifts through Berthollet, Davy, Liebig, Watt, Cartwright, etc.–gifts which have benefited him and his production immeasurably?… In a rational order which has gone beyond the division of interests as it is found with the economist, the mental element certainly belongs among the elements of production and will find its place, too, in economics among the costs of production…. A single achievement of science like James Watt’s steam-engine has brought in more for the world in the first fifty years of its existence than the world has spent on the promotion of science since the beginning of time...
As I say, it was an embarrassment—subsumed under Pivouvian externalities or discussed as requiring a tradeoff between static and dynamic efficiency via patent systems making legally excludible what was non-rival in is essence. It was a sign that Smith’s mental image of what was typical for a “commodity economy” was not describing reality, and was describing it less well as time passed.
How about information problems? They do not arise—much—in a commodity economy because you are buying and selling concrete, visible, assayable things. Neither moral hazard nor adverse selection has much play.
And problems of coordination? The image of production in a commodity economy is of a small farmer, or a small craftsman, or the work of a small team. And, where necessary, the Coasian insight—or, again, is it a dodge?—that Darwinian processes would let firms grow in size as long as their internal processes were more efficient than market coordination would have been, but then firm growth would stop, and to inquire into the workings of firms was somebody else’s business.
That is the mental image of what is typical for a commodity economy—the Weberian ideal-type. It is, of course, only an ideal-type. It was never an actual empirical reality. But it was close enough in Adam Smith’s day to guide a great deal of useful thought about economy and economic policy.
That, however, is past. What is future? What exactly, are we moving toward? Let me attempt to set out the ideal type of an attention economy, more-or-less by simply reversing the polarity on a “commodity economy” under the four headings: excludability, rivalry, transparency, and coordination.
In a commodity economy the things of value—the goods that people want—are concrete objects useful to them that are both rival (hence you “should” in some sense pay a price for using them) and excludible (hence you can straightforwardly be charged a price for access). In an attention economy the things of value—the things that you can be led to pay attention to, and would be willing in the abstract to pay for—are neither. Hence, on the one hand, you “should” in some sense have access to them for free. And, on a second hand, you should be offered great variety conforming to your taste, because it is no longer make-one use-one, it is write-once run-everywhere. But how is such a system of creators and attention-payers to be sustained when you cannot charge directly for what people find valuable in terms of what they want to pay attention to? You must, somehow, sell something ancillary that is not itself an attention good that you can somehow bundle with the attention good, and often this will take the form of a side-channel that can be manipulated to get them to direct attention elsewhere, for people who think that they can profit from it.
In an attention economy information problems are not just a complication or a distraction, they are at the essence. You cannot know what you are consuming before you do so. Once you have learned what it is, you have already paid attention to it, and your part of the economic cycle is done. This seems likely to lead to a world in which reputation and branding are of overwhelming importance: since I cannot know what it is before I make my decision to pay attention to it or not, a natural heuristic will be: is this channel one that I have found it worthwhile to pay attention to in the past, or that has a warranty of quality from something I trust?
This will, I think, lead repeatedly to the dominance of platforms and aggregators—of all kinds.
Plus there is the problem that human attention is not human utility, not to mention not human flourishing. Looking at a lot of people I know, a number of them would really be much happier and much more well-balanced if Facebook had never existed. The sharp Matthew Yglesias has, I think, a good take:
Matthew Yglesias: “Future“ & the Future of Media: ‘Content distribution… is mediated via Google… Facebook, and… lesser… smaller networks like Twitter. All… especially Facebook… deci[ded]… to give maximum visibility to content that is maximally engaging…. Engagement-maximization mostly boosts rightwing populists pushing pessimistic negative-sum worldviews while giving a secondary boost to leftwing populists pushing a different set of pessimistic negative-sum worldviews. Oops!… <https://www.slowboring.com/p/future-media>
People who are prosperous—even rich—in a material sense, and who are definitely engaged, but in a way that makes them extremely unhappy, and dangerous to their fellows who happen to look different, speak different, get labeled as different, and simply happen to be in the wrong place at the wrong time.
Then there is the question of the coordination problem in an attention economy. Back in the commodity economy, the price mechanism was an adequate communications channel because the details of the specification about exactly what was demanded were rarely if ever of the essence. Why not? Because since the division of labor was relatively shallow and simple, standard commodities could be purchased and altered to serve.
For example, consider the 30-foot bronze statue of Athene Promakhos—Athene Fighting-in-Front—designed by Pheidias that the council and people of Athens had cast and installed on the Acropolis in the year -456. 63 tons of copper and seven tons of tin went into the statue. Where were the artisans of Athens to find seven tons of tin? Herodotos of Hallicarnassos reported that nobody in Athens could tell him where the tin was coming from: all anyone could say was that the ships had picked up the tin, already mined, in Sicily, and that they thought it came from “tin islands” in the ocean. But he could find nobody who would claim to have actually seen these “tin islands”. The answer was: from Cornwall, in Britain. The market as a mechanism “knew” that 7 tons of tin needed to be mined in Cornwall and then shipped to Athens—but apparently no individual human knew that.
But as time passed, and as the division of labor advanced, coordination required much more than simply going into the marketplace somewhere, calling out your willingness-to-pay as a signal, and waiting for travelling merchants and the market mechanism to respond. In the 1930s the Ford Motor Company could run its assembly lines at full speed and efficiency only in the United States, and only near Detroit—its auto assembly-plants elsewhere struggled to attain half that productivity, even in places like the United Kingdom. And when Ford sold its dies and jigs for the Model T to the USSR when it replaced it with the Model A? In Russia, worker productivity stalled out at about ¼ the U.S. value. During World War II, from iron ore to final test at which you try to drive it out the factory, a U.S. Sherman tank took approximately 10,000 man-hours to build. A Soviet T34-C, approximately 30,000. Tight, personal, face-to-face linkages of designers to engineers to assembly workers and then on to maintenance and sales staff who have visited the factory and know what the machines can do and what their quirks are becomes essential. And coordinating this led to the modern corporations that the French christened “Fordist”.
Now, of course, the coming of the attention economy has given what Richard Baldwin calls the “second unbundling”: that of intra-firm communication. With the coming of the internet it is no longer necessary that a firm's sophisticated industrial division of labor be geographically concentrated. Faxes, datafiles, and hopping onto a transoceanic jet as a last resort: the word is that in the Before Times Apple Computer had 50 first-class seats a day back and forth between San Francisco and China. Even for those cases in which the limits of the division of labor are not so much the communication of knowledge but the face-to-face, looking-in-the-eye establishment of trust and its limits was necessary, video calls cemented by a final jet-flight-and-handshake worked remarkably well.
And so we got the international value-chain economy and the smile curve: low in the middle, high at the beginning and the end. Great value was earned at the beginning by providing raw materials and resources and, more important, industrial design. Little value was added in the middle by increasingly routinized manufacturing and assembly. And great value was added at the end by marketing, branding, and distribution—providing information (and misinformation) to consumers about what they might want from the enormous variety of types and qualities of goods that could be churned out from the expanding capacity of factories, what they would be glad that they had paid attention to. And it was, again, a quilt. Very good things happened in selected places. Other places, nearby in culture, political allegiances, and attitudes, got left behind—either their industries that they drew on for relatively high-value and high-income niches in the world’s division of good things packed up and moved away, or never arrived.
So here we have the characteristics of the “attention economy” that we are moving into in a nutshell: Increasingly, highly, highly sophisticated and integrated divisions of labor that need in no wise be geographically concentrated any more. Increasingly, information problems no longer complications but of the essence, and so we have the capture of high-value positions in the value chains by brands and platforms—and the remarkable phenomenon of clickbait in the small and the large: earning money by making your primary users more unhappy. Increasingly, the things of greatest value—with the highest notional willingness-to-pay—are experiential attention-grabbers, produced in a process of write-once, write-often, distribute-everywhere, and take advantage of what virally grabs attention to then make enough to sustain the enterprise via the sale of ancillary services.
This is the direction we are headed as we head away from the “commodity economy”. We are almost surely not going to get all the way to the end of the road. But toward this is where we are tending.
How concretely, has the SARS-COVID-19 plague accelerated this shift? They standard conventional line is that a decade of change and experimentation was packed into one year. Corporations were forced into shutdowns and social distancing, worldwide, and forced to rely on information technology to enable distributed work across a very wide range of areas. The near-consensus appears to be that companies that attempted what Steven Sinofsky calls a “skeuomorphic approach”—maintaining group meeting schedules, but doing them over zoom—failed. Instead, in his view at least, people frequently realized not that “messaging became more important or that video meetings began to work”, but that there were circumstances in which the technology-enabled form of working was actually a superior and more effective one, and so “what it really is going to be is the equivalent of what WWII did to the corporation or the microprocessor to mainframe. The question is not ‘work remotely?’, but 'what is the very structure of getting things done?’…”
As for as coordination is concerned, the tools for geographic value-chain spreading have become an order-of-magnitude better, and people’s familiarity with them has become two orders-of-magnitude better as well. Thus the centripetal forces working to undermine geographic concentration—and thus the favored places of blue-collar and other workers located near design, engineering, and management hubs—have been remarkably strengthened. So expect much more of Richard Baldwin’s “second unbundling” in the future. This will happen, not least, because it will now no longer be a reliable signal of non-seriousness for white-collar workers to decline to make the daily commute, or to move to the headquarters city.
As far as transparency is concerned, it is not clear what is happening: things could be going either way. Either people will rely more on platforms and branding as reliable signals of where they are likely to find attention-economy commodities that they should seek out, or they rely less because nearly everyone with money to spend will have spent a large chunk of the plague years building up their own personal network of reliable and trusted sources of personal information relative to them and to their preferences. As use of information technology becomes more central to those who are not aficionados—and the plague years will make it so—they will require less hand-holding, which will tend to weaken the roles of platforms and branding. To the extent that people will have learned how to consume attention-economy commodities to their great advantage—and they will have by the end of the plague years—they will demand more of what platforms are best able to provide, and have more time to spend pursuing the offerings of brands they have found reliable.
Then there is the question of attention vs. utility. I hope—but so far it has been a vain hope—that people will finally develop defenses against the Facebooks of the world, and recognize that those who want to maximize your engagement with their website so that they can sell your eyeballs to advertisers are not your friends. Human gullibility is at its maximum when we are exposed to new and strange things. But when things become familiar, a certain cynicism with respect to those who offer us things that seem too good or too surprising to be true wants to come to the fore. The plague years should help with this process. However, as I said, so far this has been a vain hope.
Last on this version of my list come rivalry and excludability. What we have seen happen to the blockbuster movie industry over the plague years tells us that it is becoming even more difficult to maintain excludability with respect to commodities that partake of an attention economy character. And as it is becoming clear to many that much of the “rival” character of commodities consumed in the old-fashioned way was not in fact of their essence, we will move further into the attention-economy world.
So what does this all mean?
I do not know. What I know is that I have talked for more than long enough.
(Remember: You can subscribe to this… weblog-like newsletter… here:
There’s a free email list. There’s a paid-subscription list with (at the moment, only a few) extras too.)
Well I think this essay is already very fine; if you can truly develop it much more, that will be impressive. If it cannot find an outlet, does that not say something bad about the attention economy?
I am not as confident as Doctorow about humans developing resistance to Facebook, because I don't think that the cases of trolling and product reviews are analogous. We learn, at root, by trial and error; and where is the feedback mechanism that would tell us that in paying attention to Facebook, we have made an error?
Demand creates supply.
We see this as far back as we can see; tool stone travels thousands of miles ten thousand years ago.
There are barriers to creating supply. The most basic is insufficient demand to balance the cost of providing the supply. (No matter how much you want an engagement ring made out of asteroid platinum, you can't generate enough demand -- spend enough money, if we resort to the abstraction -- to create the supply.)
The most common and persistent barrier is the limited minds of capitalists. Whether we point to what happened with country music once sales by genre were counted, the story of M.M. LaFleur's initial venture capital funding, or the widely observed (at least as far back as Jane Jacobs' :Cities and the Wealth of Nations:) that capital concentrations function to prevent innovation, you get supply being limited by _perceived_, rather than actual demand. And it's a narrowly specific set of perceptions.
The internet is a way to manifest the actual demand as money. Kickstarter and Patreon are obvious examples of doing this directly. (So has been a lot of sex work, and the efforts to suppress the demand are instructive about what the system actually does.)
This has three areas of consequence.
One is that the materials science to build _more_ internet or _enough_ internet spreads to the rest of materials science; lasers beget fibre optics beget better lasers and higher purity refining (because the purity of the glass determines your range between repeaters) and the techniques of this apply to all matter. (It's been at least a decade since simple commercial laser mass spectroscopy could tell you what's really evaporating off your sewage pond or going out your smokestack.)
Two is that it's not as simple as disintermediation; it's restructuring along different channels of demand. Some of the channels of demand are new because you didn't used to be able to get that at all, it wasn't possible. (Services providing personalized raunchy video; not really possible until someone invents the VCR, not really practical until you've got ~2010 internet. Small batch zirconium pen bodies; not possible at consumer prices until fifth generation composite cutting tools and post-2000 electrolytic refining. Many other examples; there's more money in video games than movies by increasingly much, there's stuff you can only do with EDM on the market, there's the price collapse of diamond abrasives, it's an extensive list.) The important thing is that the expansion of possibility creates novel demand and thus supply and this iterates. In a world where you have reviews and video and global demand, it iterates very quickly. If you don't understand the actual channels of demand -- there's pretty good evidence Patreon didn't understand their own business model! there's a lot of bafflement going around -- you're in an unstable position. People hate that, and want to make it stable.
Three is that extracting rents from your gatekeeping control of capital is now increasingly difficult. (Yes, people are making insane amounts of money. They're doing it from owning non-optional chunks of the economy, like medicine, housing, and food. If you want to make _toys_, look at what happens to physical game companies who don't understand the channels of demand. Or what happened to Tesla's valuation, if you want to make cars, and then figure out what Tesla's really supplying.) You can -- it's happened! -- figure out both what you can supply and where the real channel of demand is and do well. (Fuji's response to the death of the film camera.) Or your can try to get control of demand.
Facebook's core product is control of demand. They collect enough information to know how much demand there is for any particular thing, and they represent a sufficiently closed environment that they can hijack primate social mechanisms to create demand. It's not aggregation so much as it's hacking the signal that creates supply. (Amazon's core product is control of prices. Google's core product is control of attention. They're meta-companies in industrial commodity manufacturer terms.)
So this is not an attention economy; it's an "adjust to meaningful non-capitalist demand signal" economy. (Look at what happens when Sears gets catalog sales into the black South in the early 20th century. Same thing that's happening now globally happened then in a narrow microcosm.) And a lot of that demand is reliable social connection (pretty much all the mediated communication and mediate social activity run over the net, and thus phones) and a lot more of that demand is identity facilitation (all the fandoms). And because it's diffuse and unorganized, it's difficult to suppress. Though you can see the effort to do so, because it's recreating social order, social power, and expectations of standing.