The Downside of Imperial Collapse
When Empires or Great Powers Fall, Chaos and War Rise
Academician Victor Glushkov, the head of the Soviet program of research in cybernetics, estimated recently that, failing a radical reform in planning methods, the planning bureaucracy would grow 36-fold by 1980, requiring the services of the entire Soviet population. Such warnings are not exactly novel. Some forty years ago, the dying Lenin wrote: "Vital work we do is sinking in a dead sea of paperwork. We get sucked in by a foul bureaucratic swamp." In 1933, Leon Trotsky saw acute symptoms of the same disease. "Bureaucracy acts at random," he wrote, "it rejects objective criteria, it does not recognize laws other than the law of its own will, it substitutes commands for plans and pressure for calculation."[i]
Trotsky's indictment reads surprisingly like Premier Khrushchev's recent attacks upon Gosplan and its methods. The disease he recognized has now reached an acute stage and may seriously impede further growth of the Soviet economy. Until recently, analysis of these important processes was difficult because of the lack of information. In particular, little was known about the way in which economic decisions are actually made at the top and at the enterprise level. Interesting disclosures made recently about these previously forbidden matters have not yet received the attention they deserve.
A centrally planned economy is usually looked upon as a more or less efficient machine for the production and distribution of goods. A cybernetician would view it somewhat differently: as a machine which, more or less efficiently, generates, processes and distributes information. The two functions are intimately related. Channels of information and flows of commodities are, in fact, interrelated parts of a highly complex network. To produce a carload of, say, ball bearings, it takes not only so much steel, machinery, manpower and time; it also takes an information input in the form of data concerning the availability of resources and the demand for the product. These data are gathered, processed and forwarded to the decision-makers who issue orders to producers and receive reports which may give rise to new, modified decisions.
All these activities can obviously be handled in a variety of ways. The flow of information may take place within the confines of a small privately owned shop or, at the other extreme, it may involve thousands of messages passing among hundreds of agencies. Whatever the nature of the information system involved, a cybernetician would begin his analysis with such questions as: How much information is needed per unit of output (say, per carload of ball bearings), and what is its cost? Who receives the data and in what detail? Who processes them and by what means? Who makes the decisions? How is their execution controlled?
To provide satisfactory answers to these questions in the Soviet case and to follow our carload of ball bearings through all the recesses of the information system, nothing short of a treatise on the organization and functioning of the Soviet economy would do. Significantly enough, such a treatise would also go a long way toward answering the question how Russia is ruled. Information flows point toward centers of power. "Knowledge is power," asserted Francis Bacon.
But power wielded is not necessarily in proportion to the knowledge possessed. An acute disproportion of this kind seems to lie at the heart of Soviet planners' present difficulties. The Stalinist solution to economic planning and administration which still survives in the Soviet Union amounts to highly centralized, authoritarian decision-making, based on inadequate information and imposed upon the enterprises by long chains of bureaucratic command. To the extent that up-to-date, well-coördinated information is lacking, decisions-including some key choices which affect the course of economic development for years to come-tend to be based on rules of thumb and intuition. But intuition is a poor substitute for information in directing a complex, modern economy. Uninformed choices are costly and wasteful. The famous mathematician, L. V. Kantorovich, estimates that the introduction of more rational methods of economic planning and administration would raise Soviet industrial output by as much as 50 percent without additional inputs.
In the Soviet economy central planners decide by-and-large not only what but also how to produce. In this system, an enterprise, throughout its life cycle, is essentially a recipient of commands and a supplier of data on which subsequent commands are in turn based.
Let us examine first the birth permit of a Soviet enterprise, its so-called proiekt. Central planners have a virtual monopoly of investment decisions. Through the annual investment plan they apportion funds among industries and, within an industry, among individual projects. No industrial establishment may be legally built without a detailed proiekt, viz. engineering design, cost estimates and blueprints, prepared by a central designing organization according to the planners' directives, approved by the latter and then transmitted to the builders for execution.
One is immediately impressed by the extraordinary amount of detailed information reported to the planners and sent out by them as orders in connection with these decisions. The project of the Novo Lipetsk steel mill, for example, comprises 91 volumes totaling 70,000 pages (one is not surprised to learn that the designers are paid by the sheet . . .). "Literally everything is anticipated in these blueprints, the emplacement of each nail, lamp or washstand. Only one aspect of the project is not considered at all: its economic effectiveness."[ii] In fact, the actual cost of building a factory is difficult and often impossible to establish. "Nor is anyone really interested in it." The same is true of projects for new machinery.[iii]
Thus, paradoxically enough, while the planners receive and send out literally carloads of engineering data in connection with each important investment decision, they remain ignorant of perhaps the most crucial implications of their choices-the project's economic effectiveness and the cost of the alternatives foregone.
The basic operational directive which is to guide the Soviet factory manager in his day-by-day activities is his annual Tekhpromfinplan. This voluminous document sets out for each year in advance the volume and assortment of output, the methods of production to be employed, the cost and physical amount of each input to be used, the financial results (profits, cost reduction, etc.) to be aimed at in each quarter and month, and much more.
Preparation of that document is both time- and labor-consuming. It takes more than six months to complete the job. And then, as was recently disclosed, a curious thing happens. "This most important planning document does not perform the role assigned to it. Instead of serving the purposes of the day-by-day operational guidance, it is filed away in the factory's archives and in the Sovnarkhoz's archives and is consulted only in extreme cases."[iv]
There appear to be at least two excellent reasons for this anti-climactic action. First, despite or perhaps because of its mass of detail, the Tekhpromfinplan tends to be unrealistic and to lack consistency right from the start. Out of one thousand such documents recently examined by Soviet economists, not a single one was found to be balanced out and internally consistent. Another difficulty is that the planners are attempting, as it were, to use a snapshot, taken once a year, where a continuously working movie camera is needed. The economy, especially such a dynamic one as the Soviet, does not stand still. "Life alters plans every month, every week. Meanwhile, it takes no less than two or three months to recompute the Tekhpromfinplan of a typical machine-building factory with its 30 to 40 thousand indicators."[v]
The supply plan is another important lever of control in central planners' hands. As a general rule, Soviet enterprises are not allowed to sell producer goods to each other except by previous authorization and in amounts prescribed by higher authorities. Each summer the enterprise files with a regional Sovnarkhoz (Economic Council) a request (zaiavka) for the various supplies it will presumably need during the next year. Negotiations take place, in the course of which the original requests are usually pared down, consolidated and sent to republic ministries. In the case of some 18,000 commodities deemed to be of a national importance, final coördination and allocation take place at the central level, by Gosplan and the Sovnarkhoz U.S.S.R. The downward flow of commands takes place and eventually each enterprise is provided with buying permits and selling orders which enable it to conclude contracts with suppliers. This system was designed to prevent direct, uncoördinated dealings among the enterprises and is quite effective in that: "The buyer is separated from the seller by a long barrier of agencies and organizations. . . . For example, documents concerning distribution of tires pass through 32 echelons, ball bearings through 20 agencies. . . . Each time plans are changed, this long process is repeated."[vi] The prohibitive amount of paperwork involved may be gathered from the well-publicized experience of the Moscow automobile plant named after Likhachev. The documentation required for it to obtain its annual supply of ball bearings from the adjacent GPZ factory weighs over 400 pounds and is handled by 14 agencies.
After this laborious process is completed, the Likhachev factory will obtain the permit to buy the ball bearings it needs. But will it now get the ball bearings? This is by no means certain. In fact, its supplier, the GPZ factory, received as many as 4,000 complaints about the non-delivery or delayed delivery of ball bearings during the first six months of 1963. Persistent shortages of ball bearings have been reported throughout the country, and in some cases led to work stoppages of entire factories. The fault was attributed to errors and miscalculations in the supply plans. The same is true of various other producer goods that are centrally allocated. In the Soviet economy, which is run without adequate inventories and reserves, any such shortage tends to spread-through what might be called a "bottleneck multiplier"-to higher stages of production, where it gives rise to magnified losses. For example, due to a miscalculation in the supply of fan-belts, tens of thousands of tractors and trucks remained idle last year and grain went unharvested. The cumulative loss to the economy may thus be out of all proportion to the value of the original item and "for lack of a nail" a production battle may be lost. According to Professor V. V. Novozhilov, malfunctioning of the supply system costs the Soviet industry each year roughly 25 percent of its potential output. No wonder. How can a plant manager present detailed requests for next year's supplies and equipment before knowing what products his factory will be asked to turn out and in what amounts? Even if coördination of data at the subsequent stages of the bureaucratic process were perfect and if each supplier adhered strictly to his plan, shortages would still persist as long as the paper pyramids are erected on such shaky foundations.
Some generalization can now be made about the informational aspects of Soviet economic planning. The planners obviously attempt to obtain extremely detailed information from the enterprises and to direct their activities in equally minute detail. Professor Birman, Pro-rector of the Moscow Institute of National Economy, characterizes this approach as "the vicious attempt to gather in one document a whole ocean of commodities and to track the flows of each of them down from the center, . . . an attempt to anticipate and plan each motion a worker makes, each turn of the screw."[vii]
The system's ability to handle and digest the required data appears to be taxed to the limit. Data are collected and processed by seventeenth-century methods, complains Academician Liusternik. "Central planning agencies get literally drowned in the avalanches of documentation." And an expensive drowning it is: over ten million specialized officials are engaged in collecting and processing economic data.[viii] At the same time that central planners are "drowning in an ocean of data," they nevertheless suffer from a shortage of information. The knowledge which they ultimately distill from that ocean is largely of the wrong kind; it is useless or inadequate for making those crucial choices which, in the Soviet economy, they alone can make. It enables them to determine the emplacement of each nail in a new steel mill. But should the steel mill have been built in the first place? "Knowledge is power," but only to the extent that it renders the decision-maker aware of the various alternative courses of action open to him and of their probable implications.
From this viewpoint, the ultimate yield of the "avalanches of data" descending upon the Soviet planners is less than modest. Thus, at best, only two or three alternative variants of the annual national economic plan are considered by the planners before their final choice is made. Academician Gnedenko points out that "one cannot even talk of selecting the optimum variant" under these conditions.[ix] And its companion piece, the national supply plan, is prepared in only one or two variants.
In his important speech of February 14, 1964, Premier Khrushchev describes the way in which the key output and investment decisions were taken by Gosplan. "A circle, in geometry, is divided into 360 grades. Suppose now that these 360 parts are all apportioned among committees, ministries and departments of Gosplan, each of which takes care of only the segment allotted to it." Then, apparently, the agency in charge of "a segment" sets its output goal for the next year simply by mechanical extrapolation of its average rate of growth over recent years. Investment funds were apportioned among industries by similar means. Bureaucrats in charge of these crucial decisions often wear departmental blinkers and neglect interrelationships among sectors and products. Unbelievable as it may seem, even production of tires was planned without reference to the output of motor vehicles.
Setting targets by extrapolation obviously favors the established industries and products and impedes innovation. This conservative bias makes shifts in priorities, as set by the political leaders, difficult to implement. Khrushchev has related the difficulties encountered on that count in implementing the crash program of expansion of the chemical industry launched in I958.[x] The bureaucratic Frankenstein created in order to impose the leaders' preferences upon the economy has developed distinct preferences, momentum and inertia of its own. Trotsky's predictions are coming true.
Khrushchev's vitriolic attacks on Gosplan's methods, the far-reaching administrative reorganization of the economy following the November 1962 Plenum, the virtual scrapping of the Seven Year Plan in June 1963 and the enactment of the new Two Year Plan-all have dramatized the extent of the malfunctioning of the economy and the top leadership's recognition of the need for reform. What, then, has been prescribed to cure the ailment?
An influential group of conservatives among the Soviet planning executives and economists, while agreeing that the present system of planning does not work satisfactorily, attribute its failures to insufficient detail of information received and orders issued by the planners. They argue accordingly that the use of prices as the basis for decisions should be further restricted and replaced by ever more detailed calculations in physical terms. Decisions concerning the assortment and quality of the product which are still largely left to the managers should be taken away from them. But who would handle the huge volume of data required for this attempt to anticipate and program everything in advance? The centralizers have a ready answer: high speed electronic computers would do the job.
There is some truth in the centralizers' diagnosis but not in their prescription. Academician Nemchinov points out that since Soviet prices fail to reflect the actual cost of resources, managerial incentives based on them often work at cross-purposes and "those commodities which are in shortest supply are also often the least profitable for the enterprises to produce." But, after all, the existing prices and incentives are of the planners' own making and could be redesigned so as to assure that, to quote Professor Liberman, "what is good for the national economy would also become profitable for the enterprise."[xi] Instead, the centralizers propose to eliminate the existing feedbacks altogether and advocate an attempt to program in advance, from the center, "each motion a worker makes, each turn of the screw. . . ."
This attempt is utopian. As V. M. Glushkov has recently shown, it implies that the central planners would have to consider several quintillion relationships among the various products, probably the largest integer ever considered in economic analysis. Glushkov adds that even if high-speed electronic computers performing 30,000 operations a second were harnessed to that task, it would require one million computers working without interruption for several years.[xii] And, of course, the economy would not remain frozen, waiting for the computations to be completed.
Glushkov's picturesque illustration dramatizes the fact that the Soviet economic bureaucracy cannot be catapulted from the seventeenth into the twentieth century by the mere use of electronic computers. Academician Berg compares it to the effort to modernize a spade by attaching an electric motor. "The organizations which computers are supposed to eliminate," writes another critic, "try instead themselves to apply the computers. . . . The existing system of economic information just cannot be adapted to new conditions."[xiii]
The well-known Liberman proposals are in fact extremely moderate compared to others which have thus far passed unnoticed in the West. Liberman's central idea is: let the enterprise decide how to produce, once the planners have told it what to produce. Under this scheme, the higher authorities would continue to set output and assortment goals for each enterprise, decide who buys what products and ration out material supplies. A prohibitive amount of paperwork would still be required. To paraphrase Birman, while the Central Planning Board would no longer have to track down "each motion a worker makes," it would still have to calculate the orbit of each nut and bolt manufactured in the Soviet Union and to regulate the quintillions of relationships resulting therefrom.
Birman's own proposals for reform are perhaps the most far-reaching of all. He proposes to do away with all physical output targets except for two or three dozen key products such as steel, oil, electric power. For all the other millions of products the enterprise itself would decide what to produce so as to maximize its profits on the basis of orders received from wholesale trade and from industrial consumers. Central allocation of producer goods would be abolished. Instead, enterprises would be permitted to purchase materials and machinery directly from producers at government- fixed prices. Enterprises would compete for orders rather than evade them. Automatic self-regulation by the forces of the buyers' market would replace the present system of guaranteed sales and uncertain supplies. The enterprises' compliance with the planners' broad goals would be assured primarily by indirect regulation through prices, incentives, credit and financial policies, rather than by direct controls and communications with each manager.
An enormous saving would thus be achieved in terms of the information flows, of the amount of knowledge the planners would have to obtain and transmit. Except for the few products already mentioned, the central planners' task would be largely limited to setting the optimum proportions among the key aggregates-national income, investment, the rates of growth of the basic sectors and regions. In setting up these targets the planners would make wide use of mathematical models constructed with the aid of electronic computers.
This step from muddling through to modeling would greatly improve the effectiveness of Soviet economic planning. Instead of the present two or three variants of the national economic plan, a great many alternative courses of action could be appraised. The plan finally selected would be far more likely to assure the fulfillment of the planners' preferences at the lowest possible cost. The planners would then be dealing with the proper subjects for decision and policy. Substantial decentralization of decision-making would be achieved but, in the final analysis, while sacrificing most of their rights to receive reports and issue orders, the planners might actually have more knowledge and correspondingly more power.
How would this power be used? The focus of the debate shifts increasingly to the fundamental questions: Who is going to make the decisions? Whose preferences are the plans to implement? Until recently, discussion of these topics would have been unthinkable. Mathematical economists were going out of their way to emphasize that they intended to use their models solely in order to suggest the most economical ways of reaching the planners' goals, without questioning or analyzing the goals themselves.
But, in the light of Khrushchev's remarks upon Gosplan's targeting methods, the planners' preferences turned out to differ from and even to conflict with those of the political leadership. The implications of this state of affairs were not lost upon the mathematical economists. At a recent conference a mathematician, V. A. Volkonskii, objected to the view expressed by a conservative economist, Boiarskii, that the model-builders should take decisions on ends as given and limit themselves to the search for optimum means of reaching these predetermined ends. This would mean, Volkonskii argues, "leaving such decisions to the practitioners, without any attempt at theoretical justification."[xiv]
But what is the proper theoretical basis for making decisions on ends? Why, for example, should one product-mix be chosen in preference to other possible combinations of products? Volkonskii proposes "the democratic approach" to this key decision. The planners should ascertain what are the actual preferences of the majority of the population and follow them-an implicit reinstatement of the principle of consumer sovereignty.
The very fact that these sensitive topics can now be publicly discussed by scholars throws revealing light upon the extent of relaxation of political controls in that area. In the past, political leaders had jealously guarded their monopoly with respect to decisions on ends against any infringement, real or imagined, on the part of economists. A whole generation of brilliant pioneers of mathematical economics perished during the first Five Year Plan, having been accused of "mathematical deviation in planning," of wanting "to crank out the plan's goals" with their desk calculators. In his letter to Yaroshenko, Stalin specifically excluded any problems of economic planning from the area in which economists were allowed to roam.
But times have changed. The present generation of mathematical programmers has more to offer than the early pioneers of the twenties. Computers have taken the place of desk-calculators, highly sophisticated econometric methods have been developed. And, if economists have more to offer, leaders have now less to lose since various key decisions on goals have, in the course of time, slipped anyway into the hands of the Gosplan bureaucrats. In this predicament, the leadership may be inclined to utilize the economists' expert knowledge and to let them participate in some of the basic decisions on ends. This is implied in Khrushchev's announcement, on February 14, 1964, of the creation of a new council to be composed of scientists and economists, along with political leaders and planning practitioners, which will deal with key decisions on economic and technical progress.
What are the probable effects of the various reforms in planning methods proposed by Soviet scholars? A Western observer tends naturally to sympathize with the would-be reformers' courageous efforts, in the expectation that they would promote efficiency, rationality and freedom of choice. How justified are such expectations?
Mathematical economists believe that the substitution of programming methods for rules of thumb in Soviet planning would raise the aggregate product by at least 50 percent.[xv] Conjectural as such estimates are, the efficiency with which the Soviet system operates would undoubtedly be greatly improved. We would be well advised, however, to keep in mind that rationally allocated resources may serve highly irrational ends. Indeed, to the extent that we mistrust the Soviet ends, it might be "rational" on our part to root for the all-out centralizers ridiculed by Birman. Their attempts to plan everything would soon turn everyone into a planner. And an economy à la Glushkov where everyone is engaged in paperwork may not be efficient but is, at least, likely to be peaceful.
The crucial decisions remain, therefore, those on ends, and here the participation of Soviet scholars may have a humanizing influence. It is significant that their first venture into this long forbidden area resulted in Volkonskii's eloquent plea for respecting consumers' choices.[xvi] It would be wrong to believe that all mathematical programmers are likely to share Volkonskii's concern for democratic principles. Computers might well be used in order to invade the still remaining areas of privacy and freedom. Thus, P. P. Maslov envisions a mathematical model of the utilization of people's lives, compiled in terms of man-days of life, work, rest and self-education,[xvii] clearly an invaluable tool in the hands of a government that has never been reluctant to utilize lives for its purposes. Absolute knowledge may corrupt absolutely.
The ultimate effect which planning by computers may have on the freedom of producers' choice is as conjectural as in the case of consumer choice. Birman's vision of a partly decentralized economy where important choices are left to the plant managers' initiative may not be borne out by future events. The use of computers in American industry has, in many cases, promoted centralization. Whatever these ultimate effects may be, the current transition from charismatics to mathematics in Soviet economic planning is bound to have a profound influence upon its efficiency, upon the style of leadership and the economic climate of the country.
[i] Biulleten' opozitsii, 1933, No. 33, p. 2.
[ii] Izvestia, April 20, 1963; Ekonomicheskaia gazeta, November 10, 1962, p. 21.
[iii] T. Khachaturov, Voprosy ekonomiki, 1963, No. 11, p. 31.
[iv] Iu. Cherniak, Planovoe khoziaistvo, 1963, No. 8, p. 53.
[v] Ibid. See also Izvestia, May 30, 1963, p. 3.
[vi] N. Razumov, Voprosy ekonomiki, 1963, No. 7, p. 129.
[vii] Ekonomicheskaia gazeta, March 30, 1963, p. 7.
[viii] N. Kovalev, Voprosy ekonomiki, 1962, No. 8, p. 101; G. Pirogov, Planovoe khoziaistvo, 1963, No. 8, p. 49.
[ix] B. Gnedenko, "Pro matematichni metody v ekonomichnykh doslidzhenniakh," Ekonomika radianskoi Ukrainy, 1960, No. 4, p. 80. See also Izvestia, November 7, 1963, p. 3.
[x] N. S. Khrushchev's speeches, Izvestia, November 20, 1962; April 24, 1963; June 29, 1963.
[xi] Liberman's proposals are discussed in M. I. Goldman, "Economic Controversy in the Soviet Union," Foreign Affairs, April 1963.
[xii] V. M. Glushkov, "Ekonomika i kibernetika," Vestnik akademii nauk SSSR, 1963, No. 10, p. 11 f.
[xiii] G. Kh. Popov, "Elektronnaia sistema ekonomicheskoi informatsii," Vestnik moskovskogo universiteta, 1962, No. 5, p. 29, 21.
[xiv] Problemy optimal 'nogo planirovania, proektirovania i upravlenia proizvodstvom, Moscow, 1963, p. 502-504.
[xv] Trudy nauchnogo soveshchania po primenenii matematicheskikh metodov v ekonomicheskikh issledovaniakh i planirovanii, Vol. I, Moscow, 1961, p. 126. It is interesting to note that Novozhilov attributes at least one-half of this potential gain to the improvement in informational efficiency of planning which the use of mathematical methods would make possible.
[xvi] Another straw in the wind is the publication in Pravda (February 24 and 25, 1964) of an article by Professor A. Arzumanian, an economist, questioning the prevailing policy of a preferential growth of heavy industry.
[xvii] Voprosy filosofii, 1962, No. 3.