The Bactra Review: Occasional and eclectic book reviews by Cosma Shalizi   14

A Future for Socialism

by John E. Roemer

Harvard University Press, 1994

The Red Monday Efficient Allocation Blues

What to do the week after the Revolution has always been a poser for socialists. No doubt rallies, marches, speeches, the Internationale and the amazement of the bourgeoisie will fill the first few days, and the weekend will be no trouble, but, inevitably, Monday --- Red Monday, let us call it --- will arrive: and then what? Every sect of socialists gives its own answer, and these vary greatly in specificity and plausibility. The first ones, such as those of Fourier and Saint-Simon, were both pretty specific and pretty implausible (just between us, Fourier was a raving loon). Since then they have tended to become either less specific or more plausible. Marx devoted his energies to analyzing and attacking capitalism, and was strategically vague about its successor. Most Marxists followed this wise example, and there was even a tendency to think that there wouldn't be a problem, that by the time Red Monday arrived, what needed to be done would be obvious and simple, merely book-keeping and stock-filling. As late as September 1917, for instance, Lenin wrote like this in his pamphlet State and Revolution : "The accounting and control necessary for this ["the first phase of communist society," immediately following the overthrow of capitalism] have been simplified by capitalism to the utmost and reduced to the extraordinarily simple operations --- which any literate person can perform --- of supervising and recording, knowledge of the four rules of arithmetic, and issuing appropriate receipts."

That this bears no resemblance to what the Bolsheviks actually did on Red Monday goes without saying, and not just because Russia had a dearth of literate persons on the Bolshevik side. This vision ignores the economic problem, which is that of deciding how to use limited resources, how to make allocations. What emerged in Russia after Red Monday did, in fact, make allocations, and it's worth remembering that for a long time it seemed to the outside world to make them very well; up to the '70s, Western economists were at pains to show that, appearances to the contrary, capitalism actually did work better: or about as well: or anyhow had redeeming social values not captured in national accounts like the GNP (said national accounts, incidentally, grew out of the Soviet planning system). For quite some time many socialists regarded Soviet-style central planning as a specific and plausible solution to Red Monday.

Today, of course, one would have to search pretty hard to find someone who finds it at all satisfactory, even among those who hope Red Monday will dawn. This is not just retrospective wisdom; there were socialists and even some Marxists (the inimitable Otto Neurath, for instance) who realized that the three Rs would not be enough, and worked on the problem of how to allocate well. This was a point particularly stressed by Austrian critics of socialism, such as Ludwig von Mises, who pointed to the way self-regulating markets, setting their own prices, force people to use resources efficiently, and presented a dilemma. Without markets, the entire economy would have to be planned to achieve efficiency, which is manifestly impossible; but the market's trial-and-error method had presumably been thrown on the rubbish-heap of history the week before Red Monday. Ergo, a socialist economy cannot (except by continuing miracle) be efficient, or, as it was more provocatively put, economic rationality is only possible under capitalism.

The most interesting reply was formulated by Oskar Lange in 1936, and was to splashily impale socialism on one of the horns of the dilemma by embracing markets. He proposed to turn all allocation of consumer goods over to markets, with a Central Planning Board setting prices of producers' goods through a trial-and-error process. (He also proposed that "a statue of Professor Mises ought to occupy an honorable place in the great hall of the Ministry of Socialization or of the Central Planning Board of the socialist state.") Lange's proposal still wasn't workable --- some of the flaws will be considered below --- but it was a decided step forward, and the source of the tradition, admittedly a weak and obscure one, of market socialism.

This is the tradition of the present book and its author, who is a member of a species which is not supposed to exist, a socialist economist who is not only fully conversant with the tools of mainstream economics but uses them extensively. Economists (in America, at any rate) generally suppose that anyone familiar with their science must be loyal, if not to actually existing capitalism, then to some only slightly imaginary improvement. Such socialists as today give any thought to economic theory generally assume neo-classical economics is a pack of deliberate lies, or at best sincere but totally ideological. Nevertheless, since the late '70s Roemer has been using neo-classical tools to do highly interesting socialist economics, starting with A General Theory of Exploitation and Class (Harvard, 1982), where, having formalized and made precise some essential parts of Marxian economics, he proceeded to show that the labor theory of value and the notions of exploitation and class raised upon it self-destruct. (His proof that capitalism is impossible without the exploitation of steel is particularly delicious.) This led him to chuck the labor theory of value (saying, in his essay "Should Marxists Be Interested in Exploitation?" that it "almost always gives incorrect insights" and makes "Ptolemaic" errors), and to propose a new notion of exploitation based on game-theory and inequality in property relations. This in turn led to egalitarian political philosophy, and to the theory of mechanism design. The economic mechanism of a society is how its members make decisions, based on the information and resources available to them. The interest in the theory lies in finding mechanisms which have certain desirable properties. It includes, for instance, the principal-agent problem, that of ensuring the interests of one party (the principal) will be served by the actions of another (the agent), whose preferences may be very different. It was the combination of his studies of mechanism design with the vogue for writing socialism's obituaries in the early 1990s which prompted Roemer to write this book:

A particular way in which the modern view of capitalism suggests a future for socialism is in its understanding of the firm as a nexus of relationships between economic actors (in particular, those that economists call principal-agent relationships [cross-reference omitted]). It is not correct to characterize modern capitalist firms as instruments by which entrepreneurs capitalize their talents. The profits of firms are distributed to many owners, all or a great many of whom have no control over decisions that affect profitability and are in large part not responsible for firms' successes or failures. Firms, in other words, are run by hired agents of their owners, and this suggests that hired agents could as well run firms in a socialist economy, one in which profits would be distributed even more diffusely than they have under capitalism. Indeed, the mechanisms that have evolved (or been designed) under capitalism that enable owners to control management can be transported to a socialist framework. [p. 5]

Roemer thinks --- though this is not how he puts it --- that both competitive markets and socialism contracted mésalliances when young and easily entrapped (to unlimited private property and central planning, respectively), but that now, in their maturity, they can and should divorce these undesirables (both rather brutish creatures, really, and one, at least, more than a bit of a whore) and wed each other, and he sketches a portrait of their connubial bliss.

There is no central planning board, like the Soviet Gosplan, commanding factories to produce so many million tons of steel and size 12 leather boots. Instead, all goods, including labor (land, oddly, is not mentioned) are allocated by markets; there are many banks, making loans at interest to competing firms, who pay dividends to those who own stock in them, and are, quite cold-heartedly, allowed to go bankrupt when bad luck or stupidity force them to it. There is international trade, free in goods, somewhat restricted in capital. Market forces impel firms to efficiency and innovation, at least as much as they do today.

Clearly, a rather old-fashioned marriage, with markets as the bread-winner. Where, then, do we find the socialist better half of the union? Not in the welfare-state provisions (though there is a full set of them, like china), but in the stock market. Stock prices are quoted not in currency but in coupons, issued to citizens on attaining their majority, not convertible to cash, and reverting to the treasury at death. The price of a firm's stock in coupons will, presumably, reflect both the current value of its dividends and expectations about its future performance. Citizens can buy and sell shares in firms directly, or, more plausibly, invest in mutual funds.

Since firms cannot raise fresh capital on the stock market for investment (and there is no bond market; at least, he mentions none), they must go to the banks. Each bank has a circle of associated firms, on the model of the Japanese keiretsu, which it supervises. The terms of loans are largely market-determined, subject to some state intervention to encourage certain types of investment (and, presumably, discourage others). The banks themselves are partially owned by institutional investors (pension funds, insurance companies, etc.), and partially by the government; their boards of directors are, at least in part, democratically elected. (Roemer has a number of other tricks up his sleeve to keep banks from conspiring, like Adam Smith's merchants, with their firms or the government against the public.) There are provisions for private start-up firms, which must "go public" after reaching a certain size.

Five questions present themselves.

1. Could it work? I suspect it could; certainly Roemer's case that it could is very strong. Advanced capitalist economies are dominated by large firms, which, in practice, solve very complicated problems of mechanism design, ensuring that large numbers of disparate people act of the interest of a dispersed and largely passive set of stockholders. Roemer's system increases the dispersal of stock, but adds more mechanisms to insure efficiency. As he says, these mechanisms are already found in various capitalist countries whose economic performance, in recent decades, has been at least as good as that of the United States, which relies more on the stock market. (His discussion of Taiwanese banking in sec. 12 is particularly eye-opening.) Oskar Lange's scheme was rightly criticized by Friedrich Hayek (among others) for not giving enough reign to the market (recall that firms were to be public property, and production-good prices set by the state), but Roemer's scheme is hardly open to this objection, nor to Hayek's related one about what are now called "soft budget constraints," firms drawing on state resources rather than being forced to be efficient, or even being allowed to fail. In fact, Roemer is at times positively ghoulish about what can happen to firms in his economy.

An obvious argument against the practicality of market socialism is: "The Yugoslavs tried it, and look where it got them." Few would attribute the continuing horror there to the preceding thirty-odd years of worker-management and markets (though some writers for Z magazine seem to come close to this). Still, by the 1980s the Yugoslavian economy was in quite bad shape, and if this was due to the attempt to wed markets and socialism, that's bad news for their mutual friends. Roemer's section 10, therefore, examines "the Yugoslav experiment" and concludes that one of the supposed parties to the marriage was, in fact, an impostor:

[F]irms were not run on a competitive, profit-maximizing basis but were intensively interfered with by political authorities. Not only was competition between firms actively prevented [along with trade between the federated republics!], but the soft-budget-constraint syndrome was ubiquitous: firm managers could not be held responsible for losses, because investment decisions were often made by local political authorities. The state took responsibility for losses, covered them by printing money, and thereby unleashed inflation. [p. 86]

2. Would it be at all superior to capitalism? For an egalitarian liberal, like Roemer or myself, sure. The coupon stock market would greatly improve the lot of many people (an appendix makes some rough calculations on this head). Since the '70s, while the US national income has grown, real wages have been stagnant for most people (and actually falling for the lowest tiers); this wouldn't matter so much if wage income were supplemented by dividends. In any case, securities markets let investors hedge their economic bets, and the coupon system extends this privilege from the rich to everyone. At the same time, the evils flowing from great inequalities of wealth --- political corruption, abject and pointless want, class envy and resentment, et cetera ad condemnationem --- would be mitigated, some perhaps eliminated. (This is not to say that all social evils will be cast into the lake of fire for a thousand years --- I suspect that corruption and envy we shall always have with us --- but it takes a singularly bleak view of human nature to suppose that such things would be made worse by giving them less point.)

A related benefit, which Roemer goes into at some length, is reducing the level of profitable public bads. Many public bads, like pollution, go with increased profits, so, all else being equal, the more money you get from profits, the more of those public bads you will be willing to tolerate. If profits are highly concentrated (as today they are; in the United States, 10 percent of all households hold 83 percent of all stocks), there will be a small group of people who want much more of those public bads than the rest of the population, and, being powerful and rich, they are likely to get their way. Distributing profits widely eliminates this problem. (It may, of course, cause others.)

One of the great triumphs of neo-classical economics (already mentioned in passing) is the proof that, subject to certain conditions, markets allocate efficiently. Unfortunately, those conditions are not, and almost certainly cannot be, fulfilled in this world, and so real markets mis-use resources, to a greater or lesser degree, which is known (there is a hint of betrayal in the name) as market failure. One of those conditions is the existence of a "complete set of futures markets"; roughly, one should be able to buy insurance for any contingency. In particular, for markets to be guaranteed to work, you would have to be able to insure yourself against the risk that an investment (say, a new factory) will be unprofitable. Such insurance does not exist, and for obvious reasons (formalized as "moral hazard" and "adverse selection"), it's not the sort of thing which private insurers can offer, and consequently firms are liable to invest too little. Such insurance could, however, be organized by governments (as Roemer explains). Market socialism is not necessary before governments can take steps to correct this and other sorts of market failure --- to varying degrees they do already --- but it ought to make it easier.

3. Can it be reached from actually existing capitalism? Predicting the future is at best an entertaining party game (best played after drinking has begun), and the record of prophecy with respect to socialism is particularly risible. I will simply say that I will be greatly surprised if market socialism becomes a live option in the developed countries any time soon. At the moment, it's all we can do to preserve some fragments of the welfare state; here in the United States, we are lucky when economic policy maintains a connection to orthodox economics. But suppose we get over this, and leftists (or progressives, or what-have-you) return to strength; the key tasks would then be getting them to embrace markets, and getting the rich to give up their wealth. Roemer spends seven pages confuting "criticisms of market socialism from the left," and while nothing in them is wrong, nothing in them is very likely to change anyone's mind, either, especially not the minds of those who gravitated to left precisely because (if I may put it this way) they did not value markets at their true worth. The problem of how to expropriate the expropriators is not even dismissed.

I will be flabbergasted if market socialism takes root outside the developed countries. Roemer's version, certainly, relies on the institutions of advanced and successful capitalism. (Roemer's own hopeful remarks about China, Vietnam "or perhaps Cuba" strike me as so much whistling in the dark, notwithstanding the interesting facts he mentions about "town and village enterprises" in China.)

4. Is it socialist? Unlike Roemer, I don't find this a particularly interesting question to ask. The word "socialist" has been current for about a hundred and fifty years, in which time people have attached it, more or less plausibly, to John Stuart Mill, Joseph Stalin, fluoridation of drinking water, deliberately induced famine, keeping children and old people from starving in the streets, building tractors, building back-yard steel-furnaces, fantasies of technological abundance and leisure and fantasies of peasant medievalism and toil. My mother's family has been socialist since the 1830s, but if I were confronted by one of my Saint-Simonian ancestors, I suspect we would agree on virtually nothing. Under the circumstances, little seems to hinge on the answer.

5. How likely is Roemer's book to persuade anyone? Alas, not very. Roemer clearly wants to reach non-economists who know little math, but he has so completely assimilated the neo-classical way of thinking that the effort fails dismally. I do not just mean that terms like "market clearance" are used without explanation. Consider, for instance, the following gem, from the first section:

To put the problem slightly more formally, imagine that we can rate each possible organization of society with a triple of numbers (a, b, c) where a is the degree of equality of opportunity for self-realization and welfare, b is the degree of equality of opportunity for political influence, and c is the degree of equality of opportunity for social status with that organization. What socialists disagree about is the preference ordering over all the possible combinations: Is an organization yielding levels (2, 1, 3) better than one yielding (1, 2, 3)? [p. 14]
The reader who struggles past this to the core of the book will be treated to the detailed, careful, entirely verbal presentation of several different mathematical model economies, all of them highly abstract --- time is divided into three periods (called, imaginatively, "0", "1" and "2"), there is only one good (not even called "corn", just "the good"), etc. Theorems about these models are enunciated, again with some care, but (naturally enough, lacking algebra) not proved. The only people likely to find this persuasive will positively crave the missing formalism; I certainly did. Someone --- a political activist, say, or trade unionist --- who needs to have "equilibrium" defined will not find such models illuminating, much less compelling. (Whether they are good economics is another question.)

Some of my socialist friends regard this sort of thing as the inevitable consequence of indulging in mathematical economics (along, no doubt, with gastric pains and eventual blindness), but this is doubly unfair. First, Roemer is no worse than most of Marx, to say nothing of the suppurations of carriers of the French Disease. Second, there are mathematical economists who do write decently and persuasively --- though precious few of them. Initiates into the mysteries of the goddess Logical Rigor use a strange speech among themselves, and find it all but impossible to communicate their visions to the mass of ordinary, unilluminated mankind. This accounts in part for the fact that, of the three disciplines most devoted to that goddess, analytical philosophy and neo-classical economics have done next to nothing to shape thought and the culture at large, or even within the academy, while mathematics gave up all such pretensions long ago.

In addition to speaking Rigorese, Roemer neglects the kinds of argument which might persuade non-initiates. In softening up someone skeptical about socialism, for instance, it is essential to deal with the Soviets and their ilk. Roemer attempts to say "Why the Centrally Planned Economies Failed" in nine pages, half of them given over to an explanation (bad mechanism design: natch) which he himself says he doesn't really buy any more. His political arguments, in any idiom (egalitarian-liberal, Marxist, etc.), are few and anemic. The contrast with earlier market socialist works --- Alec Nove's The Economics of Feasible Socialism (London: George Allen & Unwin, 1983) springs to mind --- is striking and saddening.

In summary: This book makes a good case for a species of market socialism, and (here I risk prophecy) will have no influence at all.

viii + 178 pp., notes, references, index
Economics / Politics and Political Thought / Socialism, Marxism, Communism
Currently in print as a paperback, US$16.95 as of February 1997, ISBN 0-674-33946-0, LoC HX 73 R 625 [Buy from Powell's]
20 February -- 11 March 1997 (thanks to Danny Yee and Phil Apps for corrections)