INTRODUCTION
Assume there was an 'Olympiad of economic systems' with fair rules including a common starting point. Abe Bergson would be the ideal referee of such a tournament, due to his honesty and absolute objectivity. This would have been for him the dream type of competition to referee. Bergson came to study the Soviet and the Socialist system from the theoretical viewpoint of a welfare economist, as an ideal test case of his work on the social welfare function and on optimal allocation of resources. In the same way that he was not fully satisfied with a Pareto Optimum as a normative maximum, he welcomed the opportunity to investigate not only, theoretically but also empirically, in real world situations, the merits and demerits of different social and economic systems. The emerging Soviet and Socialist case provided this opportunity. That this alternative economic system happened to be dictatorial, even cruel and inhumane of the Stalin type, that it became a formidable adversary during the Cold War, were unwarranted distractions that had to be dealt with but were never the true goals of his investigation. As much as Bergson did not like that the practical implementation of any economic system involved departures from the theoretical blueprint, he reluctantly folded up his sleeves and entered the muddy field of figures, concepts and definitions in order to peel away layer after layer of real-world distortions and biases to reach the core where, hopefully the truth may be found. As it turned out, this task of peeling took away a hefty proportion of the capabilities and energy, and indeed long years of Bergson's attention3.
One may ask, and indeed some did, why such a capable theorist as Abram Bergson would spend so much time on collecting, checking, manipulating and interpreting data. There are a number of responses by Bergson. First, turning to data is preconditioned upon the prior creation of a solid theoretical structure. Bergson stated the point very forcefully in his autobiographical note: From the beginning I have been able, I think with some success, to organize my research in the light of contemporary neo-classical analysis. Should that not have been possible, I doubt that I would have persisted in working on the Soviet Economy in any substantial way (1991, p. 62).
Second, and I quote again: Data compilation can be a dreary task...But at any sophisticated level, where the concern is to relate the data to relevant analytical categories, such work can also be challenging intellectually (1991, p. 66).
A third direct quote: While university scholars formerly had to devote a substantial effort to scrutinizing official data, and where necessary making independent calculations on the socialist economies, such work has largely been taken over by the US federal government, particularly the CIA, so there is little need for academics to trouble themselves with it any longer (1991, p. 65).
Finally, the above discussion gives Bergson an opportunity to slash at those for which a... (A) corollary of the 'leave it to the CIA approach' is the inclination to employ, with minimal scrutiny, available socialist official statistics. Experience today shows that to be a treacherous procedure, (1991, pp. 65–66).
My impression is that while it is very difficult to engage so deeply and so long in the study and manipulation of data without developing some empathy for this kind of work, Bergson concluded that the search for the true verdict on the relative merit of the socialist system could not be confined to theoretical considerations alone. This belief, and also the fact that so much was written about other aspects of the work of Bergson, has directed me to dedicate this paper to Bergson's journey in search of the truth concerning socialism.
The starting point of this journey is nested in theoretical considerations, in the differentiation between the objective function of the 'system's directors' (SDs) and its implementation, and in the conclusion that implementation not only matters but is also dominant in the next section. The subsequent section discusses the evolution of Bergson's views and judgment on the merits of the Socialist system. The next section is devoted to the evolution of the study of comparative efficiency over time, and finally the conclusions are given.4
MERIT AND COMPARATIVE PRODUCTIVITY: THEORY AND/OR IMPLEMENTATION
Bergson's thinking on the comparative merit of socialism started from pure theoretical, abstract comparisons and moved gradually to the more murky comparisons of the two systems in actual practice. The emphasis is on the socialist system, though Bergson from time to time provides reminders that real-life capitalism is also not as efficient as its theoretical construct. Initially, theoretical considerations dominated, but over time, a growing number of more practical considerations, problems and distortions in the implementation of socialism were considered and emphasised as main sources of inefficiency.
Authoritarian regime and equity
The 'pure' socialist system is based on public ownership of the means of production and central planning. Its main test in a world of scarcity is the extent that it can function efficiently and therefore the efficiency of socialism was the main focus of Bergson's work. Yet, socialism as it evolved acquired two additional normative features: an authoritarian regime and a more equal distribution of income.
An authoritarian regime is strongly associated with central planning and public ownership, but is not in itself a declared goal of the system. The model of market socialism a la Lange has more freedom and more democracy. Equality and equity are clearly socialist goals but since under socialism (as opposed to 'communism') wages are determined by the amount of work done, the theoretical conflict with efficiency is muted.
Naturally, Bergson disliked and rejected dictatorship and strongly favoured democracy and freedom. The theoretical question is whether one can have central planning and public ownership without dictatorship. In his autobiographical notes, Bergson wrote: I am among those who find it difficult not to view the socialist record to date in a Hayekian way as illustrating the inherent antinomy between more or less inclusive public ownership and democratic political processes, as conceived in the West (1991, p. 67).
Similar statements are repeated throughout Bergson's work, and hence his interest in the option of market socialism where a good part of 'central' is taken away from 'planning'. This interest testifies in my view to his difficulty in coming to terms with this theoretical impossibility.5
The personal preferences of Bergson with respect to equality or equity (his 'personal social welfare function') are not that absolute but are clearly in some conflict with the declared goals of socialism: 'In the realm of equity there are, I think recognizable minimum material levels which an economically advanced society should seek to assure to all...' ..otherwise though egalitarian income distribution does not seem to me intrinsically very urgent. I recognize, however, that the degree of inequality... could affect the viability of democratic political institutions6 (1986, p. 67)
Bergson studied the issue of equity under socialism along two related lines: to what extent has income under socialism indeed been more equal than under market economies? And second, what was the impact of income equality on efficiency through possible negative effects on incentives on work and effort, investment in human capital and innovation? These two equity issues were the topic of Bergson's first investigation into both socialism and the Soviet economy. In his PhD dissertation, The Structure of Soviet Wages: A Study in Socialist Economics (1944), Bergson established first that under socialism, as distinct from communism, Marxian doctrine called for pay according to work done, basically not different than under a market system. Only during the communist stage would distribution be determined according to need (1944, 1948, 1964, pp. 7, 327–329). He then went on to estimate wage differentials in Russia and the Soviet Union and established that indeed under Stalin this principle was followed: 'despite apparent claims to the contrary, the System's directors have been able broadly to realise the principle they embrace: distribution according to work done', and possibly taking also into account work disutilities, more than under capitalism (1964, p. 329). He also observed that despite central planning and efforts to assign workers to workplaces, there has been enough labour mobility that gravitated wages toward a market-like equilibrium (Bergson, 1944; see also Millar, 2003).
Bergson continued to follow income equality and equity in the Soviet Union throughout his career. He used to complain about the paucity of reliable data. Frequently, he mentioned the special privileges of the elites as increasing inequality. In his detailed study (1984), he demonstrated with the best data available that wage differentials in the USSR may have become wider than in a number of north-western European countries, but that they were narrower than in the US and France. Bergson did recognise, of course, that in the USSR the lack of income from wealth increased overall equality. The issue of the lack of incentives for managers and inventors resulting from the public ownership and from the workings of the command economy was brought up later as part of the discussion on the efficiency of socialism (Bergson 1978, pp. 38–44; 1983, pp. 48–68).
The model and/or its implementation?
Both the political regime and the levels of equity affect the evaluation of the efficiency of the socialist system, but they should leave the field more or less value-free for the main investigation of socialism's 'pure' economic 'efficiency' or 'rationality'. The first distinction made by Bergson was between a theoretical model of socialist central planning and its practical implementation ('We must compare ideals with ideals and facts with facts', 1966b, p. 236). In his essay on 'Socialist Economics' (Ellis, ed. 1948, reprinted in 1978, pp. 193–236), Bergson supports the view, based on Barone and others, that in theory socialism can reach optimal efficiency. He goes even further and claims that the extreme statements about the non-viability of socialism in practice had already been refuted by 30 years of existence of its Soviet regime. While he had his doubts as to relative socialist efficiency in practice, he demanded that this be demonstrated by empirical investigation. Bergson seemed to reject judgment based on theoretical considerations alone (1948, reprinted in 1966b, ibid.).
The objective function of socialism
The first step in the study of comparative efficiency is the objective functions of the two systems. While this is discussed in 'Socialist Economics' (1948, reprinted in1966b, ibid.), we follow here Bergson's next major theoretical work, 'The Economics of Soviet Planning' (1964). Under capitalism, consumer welfare is the sole target of the objective function and it is achieved mainly through consumer sovereignty. Under the socialist (or Soviet) system, the objective function is determined by the 'system directors' (SDs). Bergson's first version of the objective function of the SDs under socialism includes in addition to consumer welfare the goal of rapid economic growth. He assumed that resources left after those devoted to economic growth will be directed to the maximisation of consumer welfare as under a market system (Bergson, 1964, pp. 7–8)7. It is not clear to what extent Bergson included income equality as an argument in the objective function. In a number of papers in later years he, however, mentions the possible adverse effect of equality on incentives and therefore on efficiency. There is also very little discussion concerning 'collective goods' or public services and there is no specific mention of the role of education and health; even more significantly, there is no discussion of the goal of military power as part of objective function (p. 10).
Bergson mentions two other elements that, while included in the objective function of the SDs, may negatively affect the efficient pursuit of the other objectives (1964, introduction). One such element is the adhering to Marxist teaching, especially its labour theory of value. Another is the inclusion in the objective function of the personal interests of the SDs themselves, both material and as sources of power (p. 10). This is the only mention of political economy considerations that later became an important topic in economics in general, in the study of the Soviet system in particular, and was cited by Bergson as an increasing source of system distortions.
Sources of inefficiency
While in 'Socialist Economics' (1948) most of the discussion is on the nature of the theoretical model, 'The Economics of Planning' (1964) is devoted to the description and analysis of what Bergson called the 'working arrangements' of the socialist system. The last part of the introduction (pp. 10–13) and the concluding chapter summarise the main problems and 'sources of inefficiency' that are identified in the body chapters: administration burden, distorted prices, bad success criteria, collectivisation and autarky. Under administrative burden one can include, among others, the role of information as a source of inefficiency of the socialist system. I am not sure whether at this stage the critical role of information was fully appreciated. Today one would tend to include issues of information and transaction costs as part of the model rather than as a side issue related to implementation.8 Even so the quality of implementation remains relevant and there is more than one way to collect and send information and improve its quality under central planning and public ownership, such as the model of market socialism developed mostly by Oscar Lange. Bergson seemed to see the advantages of this model, but over time he became sceptical with respect to its implementation, especially during the 1960s when reforms in this general direction were discussed in the Soviet Union and East Europe and partially implemented (Bergson, 1967). The labour theory of value represented another source of distortion, by distorting prices and therefore allocation principles and success criteria. The lack of interest on capital, and the use of average rather than marginal prices represented additional major problems. Bergson discussed bad incentives, lack of understanding, of knowledge or experience, all well known from the research of others and which do not deserve detailed discussion here.
One possible conclusion is that there is no one over-riding source of inefficiency but rather a long list of them, not necessarily interconnected. This is, in my view, an outcome of Bergson's emphasis on implementation difficulties and distortions rather than on the theoretical model. Hence, it follows that some deficiencies and distortions may be corrected and improved upon over time, like the introduction of capital charges in the 1960s, the development of computers to improve planning, better information flows and control and/or steps toward more decentralisation. In the same way, other developments may cause increased inefficiency: the main such factor discussed by Bergson was the growing complexity of the economy (pp. 354–358), a factor Bergson emphasised more forcefully as time passed (see, for example 1966, p. 240) most notably in his paper on 'technological Progress' in Bergson and Levine (1983, Chapter 2, pp. 48–49).
THE EVOLUTION OF BERGSON'S VIEWS ON THE SOCIALIST SYSTEM
By 1964, following the first empirical comparison of the relevant data for the Soviet Union and the US, Bergson (1964, pp. 340–354) reached the following conclusion on the comparative efficiency of Soviet Socialism: On purely theoretical grounds positions taken on both sides of the debate may have seemed untenably extreme....but the critics are seen to have been closer to the mark than the proponents. As exemplified by the Soviet experience, socialism must be less efficient economically than Western private enterprise (p. 355).
While cautiously taking the side of the critics, Bergson hastened to emphasise, in a special addendum to Essays in Normative Economics (1964, pp. 237–242)9 that: 'If critics have been more nearly right than proponents on socialist economic efficiency, they have not always been so for the right reasons' (p. 239). In general, this short addendum repeats arguments in 'The Economics of Soviet Planning' (1964), sometimes with even sharper criticism of the Soviet system. Then, so typically Bergson, four pages of discussion follow of potential reservations: First, he considers whether the USSR and US comparison provides enough evidence and advocates comparisons with other market and socialist countries and proposes that the valid comparison is of the two systems (models) in the same country, an impossible option. Second, he points out that socialism is 'a relatively novel economic system for which the USSR is the pioneer' (pp. 355–356). Third, after listing improvements already made and reforms that had been discussed (as of the early 1960s), Bergson concludes nevertheless, that given that socialism was nearing its 50th anniversary: 'If socialism were an especially productive system, one might think that by now this fact would have been manifest' (p. 356).
This conclusion does not prevent him yet again from repeating the discussion on the improvements in the working arrangements and restating that 'In sum, economic efficiency must already have increased in the USSR and this trend is likely to continue'. However, he hastens to add that since improvements are also expected in capitalist countries... 'One may wonder whether Soviet Socialism will ever match the latter in this respect' (pp. 356–357). Even this, however, is still not his final verdict because one has also to take into account the differences in the objective function: But whether the U.S.S.R. or elsewhere has not the inefficiency been due in part to the intense pursuit of growth? And if so would not inefficiency have been less without this goal? (And) 'The possibly favorable effect of a weaker intense concern for growth would have to be weighted in speculating about its economic efficiency (p. 357)10
Bergson also adds other particular elements of the objective function: ... 'the system's directors concerned otherwise than with economic efficiency: ideologically, politically, and personally'. (p. 356)
The final paragraph of the book is devoted to a discussion of Marx's speculation about the future. Instead of his forecast for the demise of capitalism and the rise of socialism, Bergson proposes the convergence of the two systems through corresponding changes in both (pp. 357–358).
Peter Wiles, in a review of 'The Theory of Soviet Planning' (Wiles, 1965) claimed that Bergson made a full shift of his view of the efficiency of socialism, from a positive to a negative one. Bergson denied the charge, stating that he never had a supportive view (1996a). My own thinking is that by 1964 Bergson's verdict on the relative socialist efficiency was indeed negative, but at that point he still left a window for some doubt. Or was this just a normal Bergsonian manifestation of Honest Abe?11
By the mid 1960s Bergson clearly sided with those who did not believe in a potential superior efficiency of socialism. He still, however, emphasised the uniqueness of the Soviet case, its lack of experience, the application of wrong or no economic theory, but mostly, that the observed inefficiency resulted from the particular means of practical implementation and not from the theoretical model.
Growing scepticism
Over the years that followed, Bergson became more and more sceptical, mostly with respect to the damage caused by the added elements to the objective function of the SDs on top of consumer welfare. First, he became much more outspoken about the waste involved in the goal of rapid growth. In addition, he recognised more than before the distortions caused by objectives connected with the political economy and the authoritarian regime. While he recognised the right of the SDs to have their own preferences even when private consumption issues are at stake, he emphasised more the opportunities for misuse of the powers of the SDs, their quest for power for its own sake and to serve personal and group interests. Bergson had increasing doubts as to the theoretical separability between the centralised version of socialism and the authoritarian regime (see mostly 1982, Chapter 11).
These considerations caused Bergson to (re)consider market socialism (1967, in 1982, Chapter 10; 1978). Market socialism, in its various manifestations, can ameliorate some of the above mentioned political economy weaknesses. Through its decentralised features and the use of markets, market socialism could also reduce informational burdens at the center, as the economy becomes increasingly complex (1983, Chapter 2 pp. 48–49). Not only for these reasons but also due to its decentralised nature, market socialism could also improve consumer sovereignty and welfare with a given amount of resources (ibid.). On the basis of the above considerations, Bergson's support for moves in the direction of market socialism increased: While modifying particular mechanisms, Stalin's successors have quite avoided undertaking the radical kind of planning reform... which many observers in the West once thought to be in the offing. Market Socialism also has its limitations, but the Russians may yet feel impelled to turn to that system before too long (1978a, p. 27).
In later papers Bergson also invoked equity considerations and policies as having negative impact on efficiency with respect to wages and salaries (for example 1983, p. 49) and also weak incentives for managers and innovators from the combination of pressure to fulfill output targets and low rewards for innovations. There is no direct relationship as under a market system between profits and rewards to entrepreneurship (1978, pp. 38–44; 1983, p. 59).
Finally, in his early work Bergson mentioned only in passing the burden of defense and defense as an important element in the objective function of the SDs, this factor receives more attention over time. In his 'Can the Soviet Slowdown be reversed?' (1981b) defense expenditures acquire a more important role: What remains is defense. Until now, in the face of increasing economic difficulties, the leadership has persisted with its extraordinarily costly military buildup (p. 41).
And then: The social system indisputably functioned well in a way scarcely anticipated in the official ideology, that is, in the building of military power. This is doubtless reassuring to the Soviet leadership, but it remains to be seen to what extent the USSR can exploit its military power as a surrogate for the economic superiority which supposedly was to assure the ultimate victory of socialism over its Western rival. Meanwhile the government's ability to sustain its military buildup becomes increasingly problematic in the face of mounting economic stringencies (ibid., p. 42).
Still, the burden of defense expenditures and especially that of military R&D did not play a major role in Bergson's analysis of the slowdown of Soviet economic growth.
COMPARATIVE PRODUCTIVITY, LEVELS AND GROWTH: ESTIMATES AND REFINEMENTS
The year 1964 witnessed Bergson's first in a series of attempts to directly compare economic efficiency between the two systems (1964, pp. 340–354). Thereafter, Bergson continued to pursue consecutive comparative productivity studies between the Soviet Union and other socialist countries, and the US and other Western, market economies. In addition to the expansion to more countries, made possible by the increased availability of data, the more progressive studies provided additional insights into the significance and merits of the comparisons and on the explanations and interpretations for observed differences. These discussions therefore represent the evolution of the thinking of Bergson on this main topic; they focus on Bergson's interest all along.12
The main breakthrough with respect to the availability of comparative data was provided by the European Comparative Project (ECP).13 Bergson defined the levels of efficiency or productivity as output per unit of inputs: units of labour, of capital and of 'combined inputs' (based in most cases on weights derived from a Cobb–Douglas production structure). In addition, Bergson also compared rates of growth or changes of output (GDP per capita), inputs and productivity. Bergson recognised all along that the initial comparison (1964) between the Soviet Union and the US was not satisfactory because of the wide gap in levels of economic development (LED). In his later estimates, during the 1970s and later, he followed the common practice of controlling for the LED. Bergson followed the commonly accepted assumption that productivity levels were positively correlated with LED while rates of economic growth were, based on different convergence hypotheses and the Gerschenkronian 'advantage of backwardness', to be in many cases negatively correlated with LED. Since the Soviet Union was mostly compared with more advanced countries, controlling for LED was expected to improve its relative efficiency standing but to explain away some of its growth advantage.
Bergson treated LED by comparing countries of similar levels of GDP per capita, or by controlling for two variables closely correlated with LED, the proportion of non-agricultural labour and the capital to labour ratio (K/L). In the earlier papers from the 1970s, Bergson used LED proxies as qualifiers. In later estimates, he used regression analysis. The problem was that both controls are directly related to the specific Soviet growth strategy, and therefore bias their impact. The Soviet growth strategy calls for a higher than 'normal' reliance on investment; thus, the Soviet Union tends to have a higher K/L ratio than its 'true' LED. On the other hand, the Soviet growth strategy left a higher than 'normal' proportion of labour in agriculture, biasing downward its LED. The first bias is considerably larger and therefore the combined effect of both tends to bias downwards the relative level of Soviet or Socialist efficiency. Bergson's most recent comparison was in his 1987 paper (with updates in 1992 and 1994), which used only K/L. He found the 1975 level of efficiency of the Soviet economy to be lower by 29–34 per cent than 'normal', a finding that may understate the Soviet record, considering the above-mentioned biases.
This calculation left open possible biases in the levels of output and inputs, especially of Soviet capital. Bergson used CIA estimates in $US prices. Given the various corrections made later in the levels of these variables one has to conclude that the Soviet efficiency gap was wider than described above. However even here there is an additional caveat: it is generally agreed that the conventional (CIA) estimates of the Soviet capital series were biased upward by a greater extent than the output series. If this is the case, the Soviet (and the socialist) efficiency shortfall is somewhat reduced.14
Bergson's finding of an efficiency lag of between one-quarter and one-third (for 1975 but wider in 1985 and 1990) following three generations under socialism clearly demolished any claim for superiority. The fall of the system may have been a better proof, indeed, of a more than a one-third lag in efficiency.
Bergson's conclusions on the level of comparative efficiency of socialism and capitalism, in the series of papers during the 1960s to the 1980s, provide further testimony to his caution before rushing to definite verdicts. In the early USSR to US comparisons (1971 and 1972) he invoked the excuse of being a late starter (1972b in 1978b, Chapter 5 p. 63). Later (in 1978b, Chapter 6) he stated that 'a principal moral must be simply that more research is in order' (p. 88); and, following the demonstration of the Soviet lag in efficiency, he stated: What of the comparative efficiency of socialism and capitalism generally? ... One inevitably wonders whether the relative performance .... reflects historical and cultural circumstances particular to those countries. Who knows, socialism might work better and capitalism worse if circumstances were otherwise. In short, we need further observations.... (P. 90).
Even later, when LED was already incorporated in the analysis Bergson complained that the findings were based on 'only five observations' (1971a, in 1978b, Chapter 7, p. 100), while concluding that: At least the Soviet variant of socialism seems neither colossally wasteful nor extraordinarily efficient but well within the extremes that are so familiar in polemics of socialist economics. The Soviet system, however, appears to be undistinguished by Western standards (ibid. p. 111),
He continued to ask for more observations on both sides.
Finally, on this issue I quote from a paper comparing the growth records of outputs, inputs and productivity of East (COMECON countries) and West (OECD) (1971b, reprinted in 1978b, Chapter 11 pp. 193–221): 'In the context of economic development it may well be that capitalism as it existed in the OECD, rather than socialism such as prevailed in COMECON, is the more advantageous system' (p. 218). This conclusion is reached despite of the fact that rates of growth of output in the East were somewhat higher.
Comparing rates of growth
Bergson estimated and then accepted other estimates of comparatively high rates of growth of the Soviet economy for the period of 1928 through the 1960s. These rates of growth remained comparatively high when controlled for LED. Fast growth was recognised all along as a legitimate goal of the SDs, and the high rates were explained partially as the result of initial low levels of capital and technology (see for example, 1964 pp. 354–356). At the same time too fast growth ('haste') was considered by Bergson as a waste of resources, and as one explanation for the lag in efficiency. Fast growth was achieved under the Soviet strategy, through high rates of investment at the expense of consumption and thus of consumer welfare. Efficiency suffered also from the pressure to move fast and the resulting misallocation of resources, fast depletion and thus waste of natural resources and slower technological changes. When the Soviet rates of growth started to decline during the 1960s, Bergson initially accepted this trend favourably, stating that this may reduce waste and improve resource allocation and productivity. The reforms discussed during the early 1960s were supposed, among others, to bring about an improvement (1964, pp. 357; 1966b, p. 240). This thinking may have also influenced his choice of the title of the paper mentioned on declining growth rates: 'Towards and New Growth Model' (1973). There was no question mark at the end of the title. Yet, despite the title and some of the discussion, I remember this paper as the first serious quantitative analysis of the approaching dead end for growth (and the system) under the 'extensive growth model'.
Bergson's paper demonstrated that when the capital stock grows at a faster rate than output – due to a combination of slow growth of employment and especially of productivity – the investment share of output grows and, despite continued growth, not only the share of consumption in output, but eventually also the level of consumption will start to decline.15 Since by 1970 labour reserves were nearing exhaustion (labour inputs were growing at a rate converging from above to a declining rate of population growth), the only option to reverse this trend was by substituting a few points of capital growth with those of technological change and improved productivity; hence the need for a 'new model'. Bergson ended the paper with the following: As so characterized, however, the Soviet model has clearly been undergoing a process of erosion lately. Notably... in the very sphere -- capital formation -- .. The prospect is that this erosion will continue. In the U.S.S.R at least, the Soviet model may not survive its dictatorial originator much longer (1973, p. 9).
No, this was not a prediction about the fall, except in a rather indirect way. It was an observation about what may happen. The Stalinist model, but not necessarily the socialist system was doomed. It proved to be only very partly true: yes, after 1975 (but not since 1970) the rates of growth of capital stock did decline to 6–7 per cent annually (instead of 9 per cent previously). This decline was not compensated by increased productivity; rather, total factor productivity continued to decline down to negative territory. Furthermore, defense expenditures, assumed by Bergson to stay at a constant share of GDP, grew at a faster pace (Ofer, 1987, Table 1, pp. 1778–1779; Table 3, p. 1788–1789). The changing parameters did allow some breathing space to 'muddle through' with respect to consumption, but rather than moving in the expected direction ala Bergson, the Soviet economy continued to deteriorate and thus was forced into a much more radical change.
I see in this 1973 paper by Bergson the first serious analysis of the (almost) unavoidable decline of the Soviet economy. The concluding sentence of this paper (quoted above) can also be read to mean that if the changes and reforms would not take place or would fail, the system would run into a dead end. Actually, if you read the sentence without regard to the preceding context, it can indeed be read as if it says exactly what eventually had happened.
No, Bergson, like most of us, did not predict the fall at that time or at any time later. In a number of articles during the 1970s and the early 1980s he continued to elaborate the implications of the possible introduction of a 'new model' including the changes discussed above.16 The increased readiness to cooperate with the West, including the reduction of armaments was another sign for him and others that this might been happening. While always also sceptic with respect to the outcomes of reforms that had already been implemented, there was always a discussion on what future reforms in the direction of decentralisation and market socialism might bring.17
In his autobiographical notes written back in 1986, the failure to see the fall is explicitly discussed. He also writes: 'Granting that it was not predicted, however, this abrupt denouncement is perhaps not altogether surprising in view of much Western research, including my own, briefly adumbrated here, of the malfunctioning of the economies in question' (1991, pp. 67–68). Is this an indirect reference to the 1973 paper with hindsight?
CONCLUSIONS
Starting with a model of socialism which Bergson claimed was viable as a theoretical model, Bergson spent a lifetime checking and rechecking whether the model can also pass the test of implementation. The main available test case at hand happened to be an abnormal and extremely distorted one, a case far from the scientific demand for an experimental environment with 'all other things being equal'. In addition, much of the evidence had to be collected in one of the most secretive and deceptive environments. Bergson spent a lifetime trying to isolate credible information out of mountains of false data and to create the methodology that might mimic as well as possible the required experimental environment. The title Bergson gave to his autobiographical notes was 'Recollections and Reflections of a Comparativist'. My 'Microsoft Word' spell check underlined the term with red. Bergson invented the term to best describe what was after all his academic career, indeed his adult life.
The final verdict offered by Bergson on his lifelong investigation of comparative efficiency was included in his AEA presentation in 1992. He first quoted Hayek (1935, pp. 203–204) not before noting that Hayek made this comment in 1935 at a time when capitalism was deep in recession while the Soviet Union leaped forward: It was not the possibility of planning as such which has been questioned... but the possibility of successful planning. There is no reason to expect that production would stop or that the authorities would find difficulties in using all the available resources somehow, or even that output be permanently lower than it had been before planning started. What we should anticipate is that output where the use of available resources was determined by some central authority, would be lower than if the price mechanism of a market operated freely under otherwise similar circumstances.
And then he added: 'It may not be amiss to see in my calculations of comparative productivity verification of a prescient forecast' (both p. 31).
Does this statement by Bergson refer to the theoretical impossibility of central planning to be efficient or to necessary difficulties in implementation plus special circumstances of the SU? My hunch is that over the years Bergson became more and more convinced that at least some of the factors that he classified under 'implementation' may belong to the socialist model. One such general factor is the transaction costs that are associated with central planning and that are growing as the economy becomes more and more complex. Another is the nature of the political regime associated with socialism. Over the years, Bergson observed how more and more decisions of the SDs harm economic efficiency. When the theoretical model includes also the unavoidable negative political economy implications, the model becomes deficient, not just due to bad implementation. Add to this Bergson's strong aversion of dictatorship. Playing the role of the referee in the competition between the two systems, Bergson did reach a clear verdict but did he leave the loser a benefit of a (theoretical) doubt?
Was he biased? Unfortunately Sovietology, widely defined, was infested with biased statements. From the start there were sympathizers of socialism, even communism and adversaries, both on ideological grounds. Then, during the cold war there were the political and military establishments in the West, including in the US whose interests favoured the exaggeration of the Soviet economic, technological and military capabilities. However, I wonder if we, the academic community, can claim absolute objectivity and innocence while working in this environment, and in many cases for different government agencies, which also provide budgets and data. Finally, there is the built-in bias of the 'Monday morning guessing' where so many students of the Soviet Economy as well as newcomers rushed to outdo the others and minimize the size of the Soviet economy before the fall.18 I raise this issue here only in order to declare Bergson as as honest and objective a referee as one can hope for. The only bias that may be detected in his work here and there is the quite natural feeling of some empathy that an investigator develops sometimes to the object of his life-long inquiry. I wonder what could have happened if Bergson would have titled his 1973 paper 'The Soviet Economy on a Dead-end Road?'
Is the issue of the comparative merits and demerits of the socialist or the Soviet system resolved? Is it at all relevant? Are there students who may come back to pursue these questions, possibly armed with new tools, new information and a longer perspective? During his last years Bergson lamented that there were almost no such (young) students and that the field was doomed. This may have been a premature judgment.
Notes
3 The pinacle of this effort is his monumental 1961 volume mentioned above, but work on the Soviet national income continued for many years to follow.
4 My longer paper on this subject at the American Economic Association Meetings, Philadelphia, January 7–9, 2005 also contains a section that connects the discussion of comparative efficiency to the vast work of Bergson on the estimate of the size of the Soviet economy and its growth rate; and on the debates, at the time and later with hindsight, over these estimates. The longer version can be obtained from the author and a longer exposition of the debates will reappear in a future paper.
5 Yet (or hence) in a draft of my survey article on 'Soviet Economic Growth, 1928–1985' (Ofer, 1987), I listed the authoritarian regime as one factor that contributed to faster growth in earlier stages at the future costs of transition to democracy later on. To this day I regret my consent to take this paragraph out as Bergson suggested. This was at about the same time that he prepared his autobiographical notes. I wondered for a long time what might have been the reason for this intervention. I tend to think that Bergson wanted to save me from accusation of having a political bias against the socialist system. It may have reflected a more general inclination to separate politics from economics, another support to the difficulty mentioned above. Unfortunately I never came back to discuss this with him.
6 He then goes on to invoke his origin as a son of Jewish immigrants from Tsarist Russia as influence for his preference for democracy and for the 'enduring concern for the lot of the under-privileged (ibid.)'.
7 The basic hypothesis reads as follows: 'the system's directors seek diverse goals but a cardinal concern is to assure rapid growth. Hence in choices between present and future, resource use is supposed to conform to some planners' preferences which favor the future as far as seem expedient' (ibid)p.7). 'Elsewhere, however, the concern s with 'consumer's welfare', understood essentially in terms of the preferences of households' (pp. 7–8).
8 Indeed in the discussion over this issue in Bergson, 1948 (as in 1966 pp. 233–36), he brings an argument made by Von Misses on the impossibility of the central authority to obtain the proper information in order to calculate proper prices.
9 His 1948 paper on 'Socialist Economics' was republished as Chapter 10, 'Socialist Calculation: A Further Word', pp. 237–242.
10 Harsher words on the damage of the goal of rapid growth appear in Bergson (1966b, p. 240).
11 Chapter 10 in Bergson (1966b), mentioned above, can perhaps be understood as a response to critics like Peter Wiles.
12 The main works of Bergson during the 1970s on the comparative efficiency of socialism are reprinted in Bergson, 1978, as Chapters 5–7 and 11. Later studies are mentioned below.
13 Bergson, 1991, 1995, 1997; Kravis et al, 1982; The UN, 1985, 1994; OECD, 1992; Eurostat, 1983. While the ECP estimates of the size of the socialist economies were lower in general than those provided earlier by the CIA, estimates that were based on methodologies developed by Bergson, Abe Becker and others, Bergson accepted them as a base for his comparative work.
14 In his last paper on the subject, Bergson (1994) acknowledges the possibility of such biases. He calculated that if output was inflated by 10 per cent and capital by 20 per cent, the socialist dummy would be reduced from 28 to 24 per cent. (pp. 2–3, and footnote 4). I might add that differences between alternative estimations of investment series may reflect differences in efficiency in the production of capital goods themselves and/or in their capacity to produce efficiently other goods. Therefore,in either case such biases constitute approximately a mirror image of differences in total productivity. It should be emphasised that still in the 1987 paper Soviet GDP per capita was put at 60 percent of that of the US (in US$), and the capital stock per capita at 73 per cent of that of the US! The comparative figures in rubles were of course much lower. The international comparisons, however, were always made in some kind of constant $ denomination. If GDP and capital suffer from a similar proportional upward bias, then Soviet productivity declines further. If, however, the exaggeration in the capital series is larger, then the Soviet position is improved (1994, p. 3 and footnote 4).
15 When this happens, even an autocratic regime is doomed. One illustrative calculation in this paper shows that if the capital stock continues to grow at 9 per cent annually, while the rate of growth of factor productivity proceeds at 1 per cent per annum, both as in the then recent past, consumption per capita would stay stagnant over 1970–1980 and decline thereafter (Table 2, p. 5).
16 One of Bergson's comments to my survey paper mentioned above was that I should have emphasised more the difference between the Stalin model and the one that followed.
17 See Bergson 1973, 1978b, 1981a or b and especially 1983, pp. 64–68.
18 In the eyes of a new objective function, a high proportion of the pre-fall output looks useless while many goods and services in short supply are considered of very high value. A full discussion of this issue is contained in a forthcoming paper.
