The Bactra Review: Occasional and eclectic book reviews by Cosma Shalizi   80

The Weightless World

Strategies for Managing the Digital Economy

by Diane Coyle

MIT Press, 1998

The Weightless Juggernaut, or, Prospering from the Coming Troubles

This is a book about the economic changes of the last (and probably the next) thirty-odd years, the resulting social changes, and what can and should be done about both. There are some very good, even very admirable things about this book; there are other things which are not admirable at all.

Most books on this subject are written by adherents of what we may call, for convenience, the wired ideology. The story is that an autonomous force --- ``computers'', ``telecommunications'', ``software'', ``the Internet'', or even just ``technology'' --- is rapidly and unstoppably reshaping the economy and society. The primary effect of this is to eliminate all sorts of intermediaries and all kinds of subordination. All the old rules no longer work, if they ever did; the new ones are radically different. Those who try to resist the forces of history will be swept away; those who move with it will not only reach the promised land, but become fabulously rich in the process. The future we are promised --- and can already glimpse across the Jordan --- is an anarcho-capitalist utopia, a place of flexibility, antinomianism, and unmediated communication, where the state has withered away to an irrelevance and if anything needs doing, either the Market or the Net will take care of it.

Coyle lacks the mixture of stupidity, ignorance, dishonesty and delusion needed to peddle the wired ideology, and is largely but not entirely free of its effects. This is not altogether surprising: she studied at the London School of Economics, after all, and though she's now a journalist continues to pay attention to some of the sharpest economists studying what we're doing to ourselves, like Paul Krugman, Robert Frank and Anthony Venables. Her central notion, of ``weightless'' goods, is at the very least an effective expository device, and so far as I know original with her. A brief digression will help set it in place.

It is an iron law that efficient economies concentrate their efforts on what their workers are worst at. To be less deliberately paradoxical: if a given industry's productivity (output per worker per hour) rises, then, all else being equal, fewer workers are needed to satisfy the effective demand, and so employment in that industry will fall. This will naturally tend to concentrate employment in industries which are comparatively unproductive. The prime example of this is agriculture. Pre-industrial farmers can barely feed themselves, so almost everyone must farm, and this is still the case in those countries (most of them, alas) where agriculture remains at medieval levels of ignorance, waste and inefficacy. In the country where agriculture is most efficient --- this one --- every agricultural laborer can, in effect feed, about eighty other people (neglecting our exports of food). The result is that only about 2.6 percent of our labor force works in agriculture, and that ratio continues to shrink.

The transformation of agriculture, the escape from ``the idiocy of rural life,'' is, if not the most important transformation in civilization to date, certainly in the running; but similar, smaller transformations happen in all industries. They are happening, collectively, in traditional durable-goods manufacturing, because, in the industrialized countries, productivity continues to increase in those industries, and because demand for their products (that is, ``effective demand,'' demand with dollars to go along with it) is not rising, or not rising as fast. After all, the last fifty years have seen the introduction of at most a handful of new durable goods that everyone has to have, and once a family has a fridge or a washing machine it's very hard indeed to persuade them to buy a second. These demands, in other words, can be saturated.

So manufacturing employment, as a share of the total workforce, is declining, has been declining for decades, and in all likelihood will continue to decline indefinitely, until no more of us are factory workers than are farmhands. But even in the countries most afflicted by Eurosclerosis, unemployment has not increased by anywhere near as much as manufacturing employment has declined, which raises the question of which industries now use their labor.

The answer, roughly, is that employment has shifted to two distinct sectors, information-processing and services. Neither of these are strictly new, and as I've had occasion to insist in these pages before, information-processing is, in particular, inseparable from running any kind of industrial economy, and has accordingly absorbed a large chunk of the work force since about the 1880s. It is accordingly not at all obvious that new information technologies are a truly revolutionary revolution (like the rise of industrialism), or just an ordinary, once-every-few-decades-since-1800-whether-we-need-it-or-not revolution. But both information and services are very peculiar industries, compared to agriculture or manufacturing, peculiar because what they produce is weightless.

Coyle means ``weightlessness'' very literally. Weighty goods, like a gross of pork bellies or a washing machine, are not things you'd want to drop on your foot. Weightless goods, by contrast, while they have material embodiments (even information must be physical), have them only so to speak accidentally; their value is not tied up with their mass. They are information, services, and positional goods. (A positional good is one which has value, not because of intrinsic qualities, but because not everyone has it, or if they do have some version of it, it is not one which is by common consent so nifty. This is what makes the difference between a decent shirt costing $20, or a bottle of red wine costing $12, and ones which cost ten times as much. Coyle refrains from using this term of art, but the premia people are willing to pay for style, fashion or design are largely for their positional value, and these are all acclaimed by her as prime examples of weightlessness.)

What are the consequences Coyle claims for the shift to weightlessness? Probably the most important one is the end of secure long-term, even life-time employment. I'm inclined to agree with her on the effect, but disagree as to the precise cause. Coyle seems to think that services and information are intrinsically more volatile than manufacturing, and consequently need more flexibility in hiring and firing. This seems to me to ignore the history of manufacturing firms hiring and firing at the whim of fortune before the post-war rise of the welfare states and the legitimation of unions in manufacturing.

Whether the difference between the sectors is intrinsic, or an accident of institutional history, it looks real. With it goes increasing economic inequality, the distribution of income in the weightless sectors being much more highly skewed than in the older ones. (There is a good discussion of the reasons for this in ch. 5.) The result is a simmering socio-political crisis, with three components. First, the institutions of the post-war welfare states, predicated on people either being full-time bread-winners or not working at all, fail to serve the increasing numbers who are neither. Second, political support for the inclusive welfare state continues to leach away, especially on the part of those who benefit from the higher, rising inequality. Third, public social insurance becomes even more necessary. To put the matter more strongly than Coyle does, but no more strongly than it deserves: The new economy is extremely productive, but benefits only a minority; most of the gains are captured by a tiny minority. The very large majority have, at best, stagnant incomes, and are forced to lead increasingly unpredictable, insecure lives. Something has to give; the majority will either have the system imposed upon them, or impose something in their (presumed) favor, or be given a stake in the system. The first is, arguably, what is happening in the USA (and part of why we have the largest prison-system and private police forces in the world), but it's unstable; ultimately, the enforcers will realize, like the Roman legions, the Mamluks, and many others before them, that there's no reason why they should protect other people's wealth. The second is liable to be demagogic and ineffective, if not bloody and terrible. Only the third course, the one pursued by the founders of the welfare states, offers any hope of avoiding the kind of society which the last fifty years have taught us to regard as both horrid and unnecessary. But giving the bottom eighty percent or so of industrial societies a stake in the weightless economy means, at the very least, eliminating the insecurity which currently accompanies more flexible production. This is precisely the kind of systemic, public good which any decent econ. 1 textbook shows markets cannot provide.

So far I've been discussing points where I'm in agreement with Coyle. Here we begin to diverge. Coyle recognizes the need for social welfare measures, but has some very strange ideas about how they are to be provided. She calls, for instance, of reviving the pre-welfare-state ``tradition of social provision'' of the industrial countries, though it was precisely the total failure of that tradition which ushered in the welfare states, the alternatives being either fascist or communist revolution. Again, she enthuses about local currencies. These probably help, or at least don't hurt, depressed communities, but no more than any other injection of cash, which generally stimulates productive exchange. Institutionalizing them would mean that the problems of monetary policy would be faced in every locality, and typically by people who believe in the labor theory of value and (at least in the US) autarchic, block-up-our-harbors notions about the wickedness of trade. Because she takes the ``failure of big government'' for granted, she is forced to look with favor on all manner of private, non-state enterprises --- in the wonk-speak she uses here, ``the third sector''; in English, charities --- which wouldn't begin to pass her scrutiny under ordinary circumstances. Clearly such organizations shouldn't be inhibited from trying to help people, and clearly they're better than nothing, but her view that ``if the boundary of the public sector is being forced by weightlessness to recede, the third sector will take over those areas'' is pure, unsupported optimism. The obvious objections to subjecting the poor and unfortunate to the authority of petty, free-lance bureaucrats, free of all democratic controls, elude her.

To do her justice, she does realize that welfare provisions are needed, and that charity isn't going to cut it. (It is, among other things, unreliable, and there is always an economic cost to unreliability.) For (reasonably sound) economic reasons, she favors social insurance schemes over means-tested benefits; these would also be more politically acceptable (``we're all in this together,'' etc.). Having spent three chapters or so on the Clinton-Blair orthodoxy, she admits, in one paragraph, that this is ultimately, essentially a political issue, about how we want to order our collective lives, and that its solution will be a political one, not a technical economic one.

Coyle's discussion of the general crisis of the welfare state is followed by a chapter on the looming crisis of old-age pensions --- what Americans call Social Security proper. The combined result of the lengthening life-span and the falling birth-rate is that an increasing proportion of the population is old --- for any cut-off date of ``old'' you like. In fact, what's coming up now is a one-time crunch when a truly huge proportion will be over the legal age of retirement. This means that they're entitled to public pensions, rather than having to rely on their families, or begging, or starving in the street. There are immediate fiscal problems, having to do with the way that those pensions are currently financed, which are probably surmountable, simply given political will. There are real economic problems which are not so easily swept away: there won't be enough workers in the industrial countries to support the aged in the style to which they have grown accustomed, and which they are able, politically, to demand. (This would be a problem no matter how pensions are funded, since it's about physical production, rather than simply cash, though, generally, the more cash the pension schemes have the better.)

There are essentially four ways out of this. The first is to cut pensions drastically. No doubt there are surer ways of committing political suicide, but I can't think of one off-hand. The second is to raise the age of retirement: people at sixty-five are no longer used-up wrecks, and work does at least fill the time, so this might even attract popular support. Third, immigration --- let in more people from the rest of the world who are willing to work. This would make sense --- Coyle covers the evidence that immigration is economically beneficial to all concerned in some detail --- but there are essentially only three industrial countries (the United States, Canada and Australia) where it has more than snowball's chance in Hell of being acceptable to those who are already there. Fourth, use the labor of people from poorer and younger countries, but keep them there. There's a financial problem here, too, which is that, if the industrial countries could produce enough goods to engage in this trade without running deficits, they probably wouldn't need to do so; so they'll need to acquire claims on the production of other countries now, by investing in them, and hope that those countries get rich, and won't repudiate their debts when they are. (For Europe, Japan and North America to re-conquer the world just to fund their pension schemes would be as improbable as, oh, the Soviet Union peacefully resigning its position in the Cold War as unwinnable and then just dissolving without a fight.) How this will be resolved, nobody knows.

After leaving the pensioners, so to speak, hanging, Coyle turns to the effects of the global financial markets. Government deficits --- largely, she claims, from the welfare state --- are financed through the sale of bonds. This provided the pool of capital, of securities, for the great explosion of the financial markets in the 1980s and 1990s, markets which took advantage of the new telecommunications to become electronic, global, technically sophisticated, and very fast. Debt financing also gave the markets leverage over governments. That leverage is, naturally, employed to benefit capital-holders, who tend to object to things like inflation, welfare spending, wage increases that match productivity gains, and progressive taxation. Coyle approves of this as ``vigilante justice.'' She's right that governments need some incentive to fiscal responsibility, but she provides exactly no reason to think that the testosterone- and cocaine-fueled ``Big Swinging Dicks'' of the markets are an optimal way of providing that oversight; or that they provide that oversight in a moderately reliable and cost-effective manner; or even that, on balance, that they do more good than harm. Her other argument for leaving the Big Swinging Dicks free to engorge and swing is that they are, after all, the wave of the future, ``pioneer cowboys on the frontier of the weightless economy'' (p. 180), among the vanguard of progressive mankind. --- OK, I made the last one up, but she argues (and not just in this connection) in exactly the same irresistible-forces-of-history mode as the Bolsheviks. Do they no longer teach Popper at the LSE?

Coyle's discussion of cities is better --- at first. She does a good job of explaining the (recent) economic work on why cities should exist (that is, why concentration of work and residence is economically efficient), why different cities have different economic specialties, and why we should expect all this to become more pronounced. The last point is nicely paradoxical, so let's take a minute to think about it.

Information can, nowadays, travel huge distances very quickly and very cheaply; so can certain kinds of services, which are, arguably, basically about information, like those of the aforementioned Big Swinging Dicks. This was anticipated by the pioneers of the communications revolution, who expected it would lead to decentralization, to information-processing being smeared out more or less evenly across the habitable globe. In practice, we see exactly the opposite effect, of very dense concentration in a very small number of places. Instead of having to be reasonably close to their ultimate customers (to save on communications), information-processors can now work in those places where the associated markets for specialist labor and physical supplies are well-developed, the atmosphere for the industry is favorable, and people already know that they can go there to do that kind of business. Thus these industries tend to become fixed in cities, exploiting extremely local advantages, often the results of historical accidents. Contrarily, most services are essentially non-transportable (think of nursing, or getting a haircut), and so geographically fixed. The result of increasing emphasis on activities where transport costs are either nil or prohibitive is what Krugman calls the localization of the world economy, local conditions of one sort or another becoming increasingly important, and increasingly urban.

Governments, Coyle concludes, should clearly do what they can to improve cities, as places of residence and as centers of activity. So far so good; in fact, I'm in complete agreement. But then... When she claims that the countryside has always been economically parasitic on cities, I do have to wonder what planet she's talking about. (The typical city in history isn't so much Florence, a center of production and commerce, as it is Naples, a center of ritual, administration, and the region's ruling thugs, parasitic on the countryside those thugs rule.) When she claims that the increasing economic importance of cities means they --- read, their local elites --- should be given more political authority, that, e.g., European cities should become more autonomous, like American ones, something has clearly gone very wrong with Coyle's normally acute perception of social reality. The obvious empirical counter-argument is that, as anyone who's had experience of both can attest, American cities are, on average, significantly worse places to live than European ones. This isn't decisive (perhaps, if American cities were less autonomous, they'd be even worse than they are now), but Coyle doesn't even dismiss such worries.

This is tied in to the questions of devolution and the supposed end of the nation-state, notions which Coyle treats with more reserve than the wired ideologues, but not nearly enough. It's true that many people aren't best pleased with the nation-state they currently belong to, but that almost always means that they want a different one. What, e.g., the Quebcois nationalists, or the voters for the Lombard League (and how many are there, these days?) want is the flag, the currency, the schools teaching the national language and cultural idiom, the immigration controls and the border patrol. (In less favored parts of the world: the army and the ethnic cleansing.) There's a reasonably sound instinct behind not wanting to see the nation-state wither away. Regions and cities (pace Coyle) are not large enough to provide many necessary public goods on their own, and the costs of coordinating among many independent governments of that size are prohibitive. There is also a need for an authority which has at least minimal accountability to the public, and can at least sometimes check the power of the large, authoritarian, un-accountable command economies known as corporations. Even American states, much less cities, are well-known for their spinelessness in dealing with corporations, especially ones which threaten (however incredibly) to move. The Federal government, on the other hand, does have enough power to bargain effectively with these beasts, though in the present political climate it's disinclined to use it. I'm enough of an old-fashioned liberal to believe that power should be tamed through checks and balances, so if --- as seems to be happening --- corporations are becoming so strong and so spread out that even individual states can't check them, the logical solution would seem to be more international government, rather than throwing everything back to cities, and so making the Big Swinging Dicks the sole effective arbiters of policy. As Coyle herself observes, there can be no liberalism in one country... (It might be nice if corporations were democratically accountable for their actions; it won't happen.)

No doubt it's usually true that locals know what needs doing in their area better than central authorities; Coyle, like too many others, supposes that this is enough to guarantee that devolution will increase efficiency. Alas: knowing what needs to be done doesn't imply knowing how to do it, having the capacity to do it, or wanting to do it, and all of these are necessary for effective action. Take the case of education; American primary and secondary education is the most devolved in the industrialized world, with substantial power granted even to local school-boards (which are separate from municipal governments and don't always coincide with them geographically). It is also --- now --- the worst in the industrialized world, by every quantitative measure known. As Nicholas Lemann has pointed out recently in the Atlantic Monthly, American schools are becoming more centralized and less autonomous, to pressure them into actually teaching. (Our higher education is a different story, thank the gods.)

In general, Coyle's ideas about government are the most conventional parts of her book, the most wired-ideological, and the worst.

The runner up for the weakest point is, surprisingly, the analysis of information economics. Generally speaking, free markets in information never work. The dilemma, briefly, runs as follows. Producing a new piece of information is costly, much more costly (at least since the invention of printing) than copying that information once it exists. Anyone who possess the information can, in principle, make copies of it. The result is that information will sell for what it costs to reproduce, its original producers won't recover their costs, and so less effort will go into producing information than its social utility would justify --- people only do it to the extent that they're subsidized by others, or are at least slightly crazed. Similar phenomena exist in any industry with initially increasing returns, of course, but they're particularly bad for information, because all customers are potential competitors. They get worse, because many kinds of information have ``positive network externalities'' --- the benefit to owning a particular piece of information (e.g. an operating system) increases with the number of others who also own it. Intellectual property rights are an attempt to escape from this dilemma, by allowing one party to capture the benefits (if any) from producing information. Since they are government-created and enforced monopolies, however, they suffer from all the vices which have been known to afflict such monopolies for two centuries and more, viz., the monopolists will do their best to screw the public legally. When network externalities are tossed on top of this, we get a situation highly conducive to unrestrained monopoly power, and all its customary and well-known ills. An information economy, in short, is pretty much guaranteed to experience ubiquitous market failures --- of one sort or another.

The economists who study information would find what I've just said utterly obvious. It shouldn't even be news to any inhabitant of the industrial world who's had their eyes open for the last few years. It clearly has implications for the role of governments in the economy. (Cities aren't likely to be able to do anything about global monopolies, for instance.) Coyle is trained in economics, a good journalist, and attuned to what the professional economists are saying. There is nothing about any of this in The Weightless World. Why not?

I also wish she hadn't swallowed the all-too-common line that an information-based economy is by definition clean, environmentally benign, and unlimited by natural resources. Information is always physically embodied, and there are physical limits on the energy-efficiency of computation. That's irrelevant, however, in the face of the fact that we have not so much an information economy as an electronic one, all aspects of which, from making chips and plastic boxen to sending bits over wires across the planet, are highly energy-intensive; about thirteen percent of all the electrical power in the US is currently devoted to running computers, which is an awful lot of coal being mined and burnt for something so shiny and post-industrial.

This is not as much of a flaw as it might have been, because, despite what Coyle says, the book actually has very little to do with information technology. The crisis of pensions, for instance, would only be altered by technologies producing utopian abundance or demographic catastrophe. More generally, the keys to Coyle's arguments are the joint, linked declines of the welfare state and of stable employment in manufacturing firms. That these coincided with a technological revolution in computing is more or less fortuitous, perhaps transitory. What is true of information is a fortiori true of the ``digital'' part of her subtitle. While it surely helps sales, American managers currently having a longing for digits equaled only by that of convicted Saudi burglars, none of her arguments will unwind when the next revolution comes along and we all switch to analog computers.

Let me try to cast up a total. On the credit side: Coyle has trained her sights on issues which affect all of us, and for the most part presented them well; most of her arguments are usually not just good but shrewd, and she has actually gone to the experts, rather than the witch-doctors. She writes well, and has the wit to pay attention to science fiction, popular songs and poetry. She talks sense about many controversial issues --- the relative decline of manufacturing, international trade, immigration, and the pension crisis most notably. She has a social conscience, at least a vestigial one. She realizes that markets are useful tools that let people solve certain kinds of problems, not panaceas or ends in themselves. The central notion of weightlessness is good and (I think) original.

I turn, with regret, to the debits. The largest of these, as I said, is the misguided, conventional view of government (a more accurate subtitle might've been ``reasons for not trying to manage the economy''), followed by the large gap where the information economy ought to be. She talks almost exclusively about the developed countries, with a little on the newly industrialized and nothing at all on the rest, i.e. most, of the world. She is disturbingly willing to accept poverty and insecurity for large numbers of her fellow citizens, and to consign the unfortunate to the mercies of self-appointed do-gooders. She thinks William Gibson has insights about the Internet not shared by every teenager in a chat-room. She gives no reason to think that the weightless world isn't a dismal regression from the post-war golden age, in part because she has swallowed the idea that you can't fight history hook, line and sinker.

The balance is, at least for me, in the black, but nowhere near as much as it should be.


Errata: p. 30, for ``reflecting,'' read ``reflected''; p. 72, for ``Olsen,'' read ``Olson.''
Disclaimer: I asked for, and got, a review copy of this book from the MIT Press, but I have no stake in its success.
xxii + 250 pp., black and white graphs, end-of-chapter notes, bibliography, index
Economics / the Information Society / Politics and Political Thought
Currently in print as a hardback, US$25, ISBN 0-262-03259-7, LoC HC54 C74. First published Oxford: Capstone Publishing, 1997.
10 April 1999
Thanks to Mitch Porter, Zmarak Shalizi and Danny Yee.