Printer Friendly

The Yugoslav firm versus a Wardian LMF, a socialist firm and a LMF in a capitalist market.


Ward's model of the Illyrian firm (Ward, 1958), the labor-managed firm (LMF), was inspired by the Yugoslav response to the divorce from the Soviet Bloc of the 1950s, by which central direction was to be replaced by self-management. For a long time there have been grounds for believing that, either from its very inception or since the wave of changes that started in the 1970s, the system did not differ from that of the other socialist countries as much as was trumpeted. Estrin and Uvalic reviewed the extant empirical knowledge on the Yugoslav economic system, and we can now be sure that this was indeed the case. (1) Ward's LMF model was, however, found to be of relevance in the analysis of unconventional firms and organizations in western markets: it was used inter alia to analyze the Israeli kibbutz and cooperative firms in market economies (as a recent survey by Pencavel (2012) shows). The first aim of this note is to outline the differences that did exist between the environment of the firm in the Soviet Bloc and Yugoslavia--the term Yugoslavia refers throughout this note to the entity that existed between the mid-1950s and the start of transition in the 1990s--and in other socialist economies. The second--to use the LMF model to analyze the difference in the behavior of the labor market in a labor-managed economy and in a market economy.

The Yugoslav economic model has fascinated me ever since I encountered Ward's model. Many others, I suspect, have likewise been affected. It seemed to offer the possibility of implementing market socialism, of creating a socialist market that may embody the best of all possible worlds: a world that has done away with the excessive centralization of Stalinist socialism as well as the extreme income disparities of capitalism. It may have been this promise that led to the burgeoning elaborations on the Wardian model. Yet it has also puzzled the profession for many years because there were fundamental problems with the model that served to attract a multitude of contributions, this time theoretical ones. Many dealt with the behavior of the labor market. One of its lessons was that demand for labor under a LMF regime might have the wrong slope, a theoretical property that threatened the stability of the labor market yet did not find any empirical support. Another implication was that the labor market may not clear, that it does not tend toward full-employment equilibrium and may therefore lead to enduring unemployment. The latter seemed to find a better fit with the facts, since Yugoslavia was an exporter of labor and unemployment there, unlike in other socialist countries, was widespread.

Did this literature have anything to do with Yugoslav reality? Was the Yugoslav economy really a market that coordinated the actions of socially owned, that is, non-private, competing enterprises? To understand how an economy really functions is never an easy task because the researcher is rarely privy to the deliberations of the actual decision makers. Often they make their decisions in their own minds, and even if they sit in committee and their words are recorded, one can never be sure that the pronounced arguments are what drives their deeds. Yet in the Yugoslav case, a legislated blueprint of the organization of the market lay before us. The attractions of this blueprint made many wish to believe that it described a real Illyrian landscape. But there were problems with the legislated model, and the introduction of the Basic Organizations of Associated Labor, the BOALs, in the 1970s, was the most serious one. The BOALs system seemed to split up all social organizations, even the labor force of an enterprise, into countervailing monopolies that had to arrive at social agreements for the united implementation of a production and distribution plan (see Prasnikar and Svejnar, 1988). This arrangement, which ostensibly enabled each part of an economic unit to hold it up, seemed, from its inception, not to be feasible and thus strengthened the question mark hanging over our understanding of how the system functioned in reality. (2)

The reasons we need to understand how the Yugoslav system worked extend far beyond the confines of the country that was Yugoslavia. They touch upon the wider question of the nature and viability of a socialist market. If Yugoslavia was indeed a socialist market, then the question of viability is solved, though not of its efficiency. If it was not, then the doubts regarding the feasibility of such a market (eg, Keren, 1993) deepen, although a single instance of a system not working does not prove that it cannot function. Well, was the Yugoslav economy a socialist market? And did the LMF model apply there?

Many authors supplied partial answers to this riddle, among them Uvalic (particularly 1992) and Estrin (1983 [2010]). Estrin and Uvalic's review (2008), which covers the extant literature, indicates that the labor-managed economy was really a very thin veneer that was hiding real economic relations that did in fact not differ much from those of the other socialist countries. The system was not managed by its decentralized labor units but was largely managed by outside authorities, with the lower rungs of the hierarchical party, the League of Communists of Yugoslavia, serving as the guiding organ. (3) This explains the conundrum of the BOALs, the fact that a system that purportedly delegates coordination to a myriad of countervailing monopolies, does nevertheless manage to function. It was not decisions by these monopolists, the individual labor councils or BOALs, that determined the policies of the Yugoslav firm (YF) and coordinated flows in the economy, but decisions made by the republic party organs. As in traditional socialist economies, allocation and resource use decisions were guided by political organs. This was their main finding. Not enough is known about the exact administrative structure of the allocation mechanism (but see Prasnikar and Svejnar, 1988). The type of detailed information that the post-War Harvard Project provided (Berliner, 1957, 1959; Granick, 1972; Levine, 1959,1961) does, to the best of my knowledge, not exist for Yugoslavia.

A point that needs stressing is that it was the state or republic level League of Communists that made economic decisions and that local administrative bodies, too, were deeply involved. (4) Yugoslavia, especially in its last decades, was a real federal state, and the League of Communists of each of the republics rather than the federal party ran the economic show in its own manner (Prasnikar and Svejnar, 1988). Thus party bodies inside, say, Serbia, Slovenia, and Bosnia-Herzegovina ran the resource allocation organs of their respective states, but interstate economic flows were often arranged by market forces. This also meant that YFs had market experience that was lacking in the orthodox socialist economies, though not in reforming countries such as Hungary.

Price-taking competition between LMFs did not exist, yet workers' incomes were affected by their enterprise's profits, in spite of wide-scale redistribution of profits, because incomes in different enterprises varied widely. And employment decisions must have been influenced by these workers' interests. Furthermore, the political and economic systems of Yugoslavia were endowed with more popularity than their equivalents in other communist countries. This goes back to manner in which the regime was imposed after WW2--by local forces rather than through an external power--and to the manner in which the break with the Cominform was effected, both of which inspired pride in the regime. The labor management model itself, with the sense of worker control that it imparted, had more appeal than the socialist system in other communist countries. This meant that YFs were more difficult to change and privatize from above than state-owned firms of the other socialist states. This must have affected the path of transition (Uvalic, 2010).

Why come back and reexamine a system that has passed away over two decades ago? What can this paper add to its story? The answer to the first question can be found in the opening paragraph of this paper. Since the socialist market is for many a very attractive solution to many of society's ills, especially at a time when trust in capitalism is being eroded by the 'worst recession since the depression', it is important to establish whether it is at all feasible and how it may function. Furthermore, the question per se is of interest. And the labor management model has relevance for cooperatives in western markets, as evidenced in Pencavel's enlightening recent survey (2012). (5)

The aim of this note is to extract the principal characteristics, stylized facts, of the YF that may help amend the LMF model. These cannot always be true to all finer details, which, as Prasnikar and Svejnar (1988) recount, were not uniform throughout the country. It focuses on the institutional and behavioral aspects of the environment that distinguish the YF and Yugoslav market from the traditional socialist enterprise on the one hand ('The Yugoslav and the socialist economies--A brief comparison') and from the cooperative in a market economy on the other. Unlike the accepted treatments of the system, it stresses the fact that labor cannot be considered a variable factor, neither in the YF nor in a cooperative. It also uses a graphical analysis that is novel. It then restates the traditional LMF model in the section 'The traditional model' with slight modifications, and shows in the section The model versus the YF: BOALs and centralization, administrative allocation and wage determination' why the outcome of the ostensibly decentralizing BOAL reform was centralization. In the subsection 'The real LMF: No firing' it concludes that labor is in effect a fixed factor. In the next section it compares the functioning of the LMF in a capitalist and a Yugoslav-type labor-managed market, explaining why, unlike in the latter, in the former the market is equilibrating. The final section concludes.


Yugoslavia started its socialist era as a centralized Stalinist economy, like the other Soviet Bloc countries. Its groping transition to a more decentralized system was perforce gradual. Nevertheless, three distinct periods can be discerned in its history. (7) During the first period, which started with Yugoslavia's ejection from the communist international, the Cominform, in 1948, and lasted until 1965, enterprises had little latitude in exercising their purported freedom, because most of the critical economic functions, for example, capital allocation, remained in central control. The second period, 1965-1972 was the most decentralized and closest to the Wardian model. It had many attributes of market socialism, retaining social (non-private) ownership in the bulk of the economy, with a unique system of workers' self-management that ensured formal mechanisms for workers' participation in (at least some) enterprise decisions. (8) The reaction came in the 1970s with the BOALs reform that was designed to make self-management more participatory and democratic by bringing decision-making forums closer to the worker and to equalize labor income. Yet it did this by splitting up each enterprise into small interdependent work groups, which, like all other organizations in the economy and polity at large had to agree on interlocking contracts. This unwieldy system needed external decision makers to coordinate. Thus an ostensible attempt at decentralization led to effective recentralization (see the section 'The model versus the YF: BOALs and centralization, administrative allocation and wage determination').

One of the characteristics of the Yugoslav economy has been unemployment, a phenomenon foreign to the other socialist economies, where labor, like all inputs, was hoarded. Table 1 summarizes the statistics provided by Woodward (1995). Two facts stand out: one is the increasing level of unemployment, mainly after 1975, and the strong relation between development and employment: Slovenia was fully employed throughout the period. Another characteristic was a continuous rise in income dispersal across regions and industries, and even among workers of similar skills in different industries and different firms in the same industry. (9) This is what the model predicts ('The traditional model').

The YF, in spite of this history, in many respects differed little from the socialist enterprise. Self-management in the YF was highly constrained and its wage policy and, to some extent employment, were strongly circumscribed by the League of Communist's local and republican authorities. In this, it was similar to the Hungarian enterprise of the post-1972 stage of the New Economic Mechanism (Portes, 1978a). Like all socialist enterprises it too had a soft budget constraint: YFs were rarely allowed to go bankrupt, because its missing cash flow would always be made up by various means--bank loans (later rescheduled or forgiven), transfers of profits from other YFs, credits by other YFs, or a merger with another YF. (10)

Yet the differences were significant. First comes the federal structure of the country, which meant that there was no overall coordinating authority, no industrial ministries, no Gosplan. Yugoslavia was essentially an open economy, and there was no need to balance inter-enterprise flows. Although the policy of creating large monopolistic producers was similar to that of other socialist countries (Estrin, 1983 [2010], p. 82), federalism made a difference and allowed competing firms to be established in other republics. The YF functioned therefore in a kind of a market, but its soft budget constraint meant that the strong incentives for efficiency that a competitive capitalist firm experiences were missing.

There was also a difference in the behavior of the YF. The fiction of the LMF did have an influence, and the membership felt that it was at least part owner of the firm and claimant to its profits. Furthermore, the YF had no state plans imposed on it and no targets to reach. It was therefore in no need to hoard inordinate quantities of produced and primary resources. In particular, there was no need for excess hoarding of labor, a common feature of socialist economies. The YF could influence if not determine the size of its membership (see the section 'The real LMF: No firing' below). And since profits did affect members' incomes, if not determine them, members were minded to keep them to themselves, that is, not increase membership beyond the size that would in the long run add to their individual incomes. (11)

For reasons elucidated in the section 'The real LMF: No firing' below, the LMF does not consider labor, that is, membership, to be a short-term resource, easily adjusted when conditions, for example, prices change. The size of the workforce is determined as a long-run basic decision in step with the determination of the desired stock of capital, and short-run price changes do not lead to changes in membership. Hence the specter of instability in the labor market is not one that requires serious attention.

Finally--transition. On two counts transition through privatization is more arduous for LMFs. First comes the legal problem. State-owned property can be sold by the state. Not so social property that ostensibly does not belong to it. If the distinction is taken seriously, and it would be a denial of a history of over half a century not to, social property has to be renationalized before the state can sell it. Next, and more complex--the politics of selling off property that workers consider (at least partly) their own. This feeling that the labor management model is more than a fiction, strengthened by the pride in Yugoslavia's own way and the fact that communism there was not imposed by foreign powers, meant that workers had to be bought out more dearly than in state socialism countries. That this was indeed the case is shown in Uvalic and Vaughan-Whitehead (1997), where the legal process of privatization in socialist European countries is detailed, compared with Uvalic (1997), where the privatization stories of the various splinters of what was Yugoslavia are told.


The traditional model

The basics of the labor management model are well-known and can therefore be presented very cursorily. The YF, a cooperative owned and managed by all its workers-members, produces the output, Q, by its members' labor, L, consisting of equally productive members of the firm, and other fixed factors whose total cost is FC. FC comprise capital costs as well as taxes, assumed here to be idiosyncratic, firm specific, and given. Thus

Q = Q(L) (1)

where Q is locally convex, with Q" < 0 and Q' > (Q/L) for low L, Q' <[Q/L) for high L. Each member's consumption, her maximand y, consists of wages, w, plus per-member profits, which equals total cash flow, that is, the value of output, whose price is p, minus fixed costs, per member,

y = w + [p.sub.Q](L) - WL - FC/L = pQ(L) - FC/L (2)

which is unrelated to the level of wages. This is maximized when

P[Q.sub.L] = y (3)

(again, regardless of the wage level and unaffected by it). This is the conventional representation, or

Q(L)/L = [O.sub.L] + FC/p/L (4)

The latter highlights the utility of new members to the existing ones: it consists in the reduction of the burden of real fixed costs. This way of putting it also shows why the 'demand' for new members has the 'wrong slope': an increase in the price of output reduces the real burden of FC and therefore the contribution of new members. This is illustrated in Figure 1, where the horizontal axis denotes the membership, L, and the vertical--the real wage or income in real dollars ($/p, the $ sign is used for convenience) per member or worker. The broken Q/L is labor's average physical product and the thin MPL is its marginal product. The dotted FC/pL is real fixed costs per worker-member, ie, the monetary costs normalized by the price of output. The real wage, [wp.sup.-1] declines with p: the dotted curves measure FC at wage-labor units. There are two such lines, a black dotted FC/[p.sub.0]L and a gray dotted FC/[p.sub.1]L, where [p.sub.0]>[p.sub.1]. The member's income or consumption is the thick curve, y, where y = (Q/L) - (FC/pL). Since there are two FC/pL lines, there are also two y lines, a black one for [p.sub.0] and a gray one for [p.sub.1]. The latter lies above the former because [p.sub.0]>[p.sup.-1.sub.1]. The highest LMF's income-per-member is the point of equality of MPL and y, and we have two such points, [L.sup.*.sub.0] and [L.sup.*.sub.1]. [L.sup.*]([p.sup.-1.sub.0]) <[L.sup.*]([p.sup.-1.sub.1]) even though p.sup.-1.sub.0] < [p.sup.-1.sub.1] and the latter results in higher employment and membership coupled with lower real income ([y.sup.*.sub.1] < [y.sup.*.sub.0]) See Appendix A. This is the problem of the wrong slope of goods supply and--what concerns us here--the demand for membership-labor. The wrong, that is, positive relation between real wages and the demand for labor, is a problem only when it is more elastic than labor supply, because then excess demand (supply) for labor will increase in response to a rise (decline) in wages, moving the market farther away from equilibrium and making the market unstable. Any reduction in the elasticity of demand will therefore reduce the likelihood of such a condition. The following sections argue why this this elasticity is small ('The model versus the YF: BOALs and centralization, administrative allocation and wage determination') or even zero ('The real LMF: No firing'), and this is a non-problem.


The model versus the YF: BOALs and centralization, administrative allocation and wage determination

The ostensible aim of the BOAL reform was to democratize the economic sphere by decentralizing enterprise management, that is, by splitting the YF into smaller work units. One may think of it as a multi-divisional labor managed organization, an M-Form LMF. The manner in which this was done differed diametrically from the manner in which GM, the prototypical M-Form, was split up into its divisions, each of which was capable of functioning independently. In Yugoslavia it was done by splitting establishments up into sub-units, each of which was dependent on all others and incapable of producing anything by itself. All individual BOALs belonging to an interdependent enterprise have equal standing, and all together are to arrive at an agreement, a contract specifying the production plan and the distribution of the surplus, by negotiating among themselves (Prasnikar and Svejnar, 1988). Decision making, in other words, was delegated to clusters of units, each of which has a veto power, each of which could force a stalemate. Such an unwieldy system was manifestly unworkable. And, indeed, it had a super structure of social and regional organizations--all ruled by the various levels of the hierarchical League of Communists--with which each of firm had to sign social compacts, and these enable the economy to function, that is, by decisions of these higher bodies. As various authors (eg, Prasnikar and Svejnar, 1988; Vodopivec, 1993; Estrin and Uvalic, 2008) assure us, decisions were imposed upon the YF from above, although the loci of power differed across industries and regions (Prasnikar and Svejnar, 1988, pp. 241-242). In other words, malfunctioning decentralization led to centralization. The pre-BOAL system was, by the evidence we have, fairly decentralized (Estrin, 1983 [2010], pp. 66-72).

In other words, the BOALs reform led to increased centralization, which affects the very foundations of the Wardian model. It means that the production and employment decisions of the YF, at least since these reforms, were not self-labor managed, but directed from outside, that is, by local League of Communists authorities who imposed their policies on the management. Prasnikar and Svejnar (1988) and Vodopivec (1993) report that wages for particular groups of workers were not fixed by means of negotiations among these groups or BOALs but were adjusted by the management, subject to constraints imposed by republic and local party organs. (12) But what were the criteria guiding these organs? What bureaucrats usually do in similar situations is to form their decisions in the following manner:

* Start out from the wage structure inherited from the past;

* Take account of the relevant principles, that is, the labor management concept that the enterprise profits in some sense belong to its members, and adjust all wages alike in some relation to profits;

* Give consideration to ad hoc governmental instructions.

* Be influenced by idiosyncratic events, such as special demands of some groups of workers or of labor market pressures, for example, the scarcity of certain skills.

Assuming the latter two to be random, we can take it that incomes were corrected each year by some small proportion of profits. In other words, (2) is replaced by

[y = y.sub.-1] + [alpha] (pQ(L) - [y.sub.-1] [L - FC/L) = (1 - [alpha]) [y.sub.-1] + [alpha] (pQ(L) - FC/L) (5)

and the optimum membership, defined by (3) or (4), is adjusted over time (see Appendix B):

y = [apQ.sub.L] + (1 - [alpha])[y.sub.-1] (6)

In other words, (13) the optimum membership, though unchanged, may over time be slowly adjusted when fundamentals, say, prices, are seen to change permanently: it is now defined as a weighted mean of the existing L and the long-run LMF membership, where the weight of the latter is [alpha], a small number. Thus the threat of a high positive elasticity of labor demand recedes--its slope is between the perpendicular fixed L([y.sub.-1]) and [pQ.sub.L] that is rising with p, with a very small weight to the latter. Yet see the next section.

The real LMF: No firing

The assumption of the traditional model is that the LMF may choose at any point of time freely to hire and fire its members. Or that it hires anew its members each period. These assumptions can surely not be applied to the real Yugoslavia--nor to any real cooperative--as has been pointed out by Robinson (1967), Vanek (1969), Meade (1972), and others. We have to assume that the reduction of membership, that is, of the labor force in a cooperative community such as a LMF, is very costly if not impossible. This is a significant change, because Ward's original assumption leads to the result that as p rises and with it the real income of members, demand for labor falls. That is, when profitability rises and improves the condition of the members, the latter get rid of some of their mates to raise even further the incomes of the remainder. This is what threatens stability in the theoretical labor markets of the labor-managed economy, yet it runs counter to fairness and intuition.

The simplest way to amend this untenable assumption is to assume that the labor force can be increased but not reduced. Suppose we take the price p to be the parameter that may change, with yesterday's price [p.sub.0] and today's [p.sub.1], declining by [pi] to [p.sub.1] = [p.sub.0] - [pi], [pi] > 0, raising the burden of FC and hence the contribution of additional workers just enough to justify the addition of a single worker for today. Yet if we add this worker to our LMF, we are stuck with her for life, and when prices in the future rise, by hiring her we shall have burnt the option of staying at our current size. If the likelihood of future price rise is significant, the small gain today will not justify forgoing the option of staying slimmer tomorrow. As Prasnikar and Svejnar (1988) state, 'the fixity of enterprise employment ... makes firms cautious about admitting new worker-members who cannot be easily laid off later' (p. 254). As a result, the demand for labor will be very inelastic and the theoretical rise in demand for labor as prices fall (and real incomes rise) is all but eliminated. This is not hard to model, but the result seems too obvious to justify this further encumbrance.

The LMF model has to adapt to the very dear, perhaps prohibitively dear cost of reducing membership. The simplest adjustment is to assume that labor in the LMF is as fixed as capital, that is, that membership is determined for the long run, on the basis of expected future market conditions, prices included eliminating labor demand elasticity and the danger to stability. This assumption is maintained in what follows.


Figure 2 is the basic tool used to analyze a local labor market. Let us start analyzing cooperatives entering a local capitalist labor market with an established competitive wage level of [w.sup.*] (Figure 2). The total supply of labor in the local market is [O.sub.a][O.sub.b]. Two cooperatives enter the market: the curves representing Cooperative a are those of Figure 1, except that Q/[L.sub.a], which is just a reference curve not used in the analysis, is drawn in gray. The membership of Cooperative a is [O.sub.a][L.sub.a], and each member's income is [y.sup.*.sub.a] (see left axis). The curves of Cooperative b are also based on Figure 1, but this time flipped sideways to the right and drawn as broken black or gray curves, respectively. Its membership is given by [O.sub.b][L.sub.b] and members' income by y*b on the right-hand axis. The remaining workers, LaLb, are employed by capitalist firms at the wage w*. Observe that each cooperative determines the size of its membership without regard to the market wage, except that we may believe that most workers will choose not to join cooperative j if [y.sub.j]<[w.sup.*], j = a, b.


The same figure, Figure 2, is also used to analyze the labor market in the labor managed economy. The total supply of labor in the local market is again [O.sub.a][O.sub.b]. This time there are two LMFs in the market: the membership of LMFa is OaLa, and each member's income is [y.sub.a] (see left axis). The curves of LMFb are again flipped rightwards, and its membership is given by [O.sub.b][L.sub.b] whose income is [y.sub.b] on the right-hand axis. This time [L.sub.a][L.sub.b] represents unemployed workers: each LMF determines its demand for membership without regard to the market. Unlike a capitalist firm, to which a newly hired worker costs his market-determined wage, the cost of a new member to the LMF equals the income of existing members, and is not at all affected by market conditions. Wage changes do not influence it. There is thus no equilibrating mechanism in the labor-managed market, except the entry of new firms, that may lead it to full employment. But decentralized entry is problematic because of constraints on the capital market: a capital-less group of unemployed workers will not be granted loans, because, having no own capital to endanger, their incentives will be to invest in high risk projects (Stiglitz and Weiss, 1981).


Using the same figure, a model of a capitalist market economy does produce a full-employment equilibrium allocation, at the meeting point of [MPL.sub.a] and [MPL.sub.b], of [O.sub.a][L.sub.e] to firm a and [O.sub.b][L.sub.e] to firm b, with a wage of [w.sup.*]. Observe that in this case [y.sub.a]>[y.sub.b]>[w.sub.t], though it is easy to show cases where the capitalist wage is in between the two LMFs' members' incomes. It is true that unemployment may exist in capitalist markets in spite of what the model shows. Yet the worrying fact is that in the labor managed market there are no market clearing forces, forces that drive it toward full employment.

This statement needs qualification: There is one case in which the labor managed market will be fully employed--see Figure 3. Here the real fixed costs of both LMFs have been increased, raising the demand for membership until supply is exhausted. This case could represent a capital-endowed market, in which many high-FC firms compete for the available labor. This could possibly represent the Slovenian labor managed market. (14) Thus the combined income maximizing memberships in both LMFs, [O.sub.a][L.sup.*.sub.a] for [LMF.sub.a] and [O.sub.b][L.sub.b] for [LMF.sub.b], exceed available labor, [O.sub.a][O.sub.b]. Will the resulting income for each LMF member in this market now be equal, with [LMF.sub.a]'s membership [O.sub.a][L.sub.e] and that of [LMF.sub.b] [O.sub.b][L.sub.e]? This would be the case had the assumption that FC were firm-idiosyncratic been given up. In that case, the MPL of both LMFs would be equal and

general equilibrium solution in the labor-managed-market would be identical to that of the capitalist market, the case analyzed by Meade (1972) and Vanek (1970). Yet given our assumption that the second input is not fungible capital but idiosyncratic fixed costs, specific to each firm, the higher income firm, [LMF.sub.a] would keep its preferred membership at [O.sub.a][L.sub.a.sup.*], because no member would be induced to leave it for lower income [LMF.sub.b]. And the latter would be short of members at [O.sub.b][L.sub.a], which is less than [O.sub.b][L.sub.e] and much less than [O.sub.b][L.sub.b.sup.*]. Even where both firms' incomes per member were equal, considerations of stability would lead one of the firms to achieve its optimum membership while its competitor would make do with the remainder. Observe that a situation like LMFb's, with employment at a point where MPL is increasing and second order conditions do not hold, is tenable for a LMF, because its members have no market wage to accept if they leave their firm, but have to do with their firm's mean income.

To sum up, what the simple model teaches is that, in the labor-managed economy, there is no labor market force that would drive wages to any full-employment equilibrium, and there is no tendency to any equalization of wages across workers in any location or of similar occupations and skill across industries and locations. That this result of the traditional model is relevant we learn from reports of high levels of chronic unemployment in Yugoslavia (Table 1) and large and growing spreads between income levels of workers in the same location, even those having the same skills (Estrin, 1983 [2010], Chapter 5 and passim).


An economic system is a notoriously elusive animal to analyze, because the nature of the transactions going on and their actors' motives are hidden and very hard to decipher. We should therefore be particularly grateful to the examination of the labor market by Estrin (1983 [2010]) and the capital market by Uvalic (1992), and to their joint final definitive paper for demystifying and clarifying the conundrum of the Yugoslav economy. Now, nearly two decades after the disappearance of the Yugoslav economic system, we possess a fairly clear picture of its operation. The aim of this note was to disseminate succinctly their main findings, extract simple generalizations that can be applied to the model and, where possible, to supply a simple diagrammatic explanation to the microeconomic foundations for the system which they describe. An additional aim was to see where the old LMF model, once it can no longer be said to present a blanket depiction of Yugoslav micro-economics, is still applicable. We find that it is applicable to some aspects of the Yugoslav market and to the behavior of cooperative firms in the capitalist world.

The paper argues that the main bugbear created by the model, the danger of market instability, has no roots. If the demand for labor declines when real wages fall, as the Wardian model predicts, and if labor demand is more (positively) elastic to wages than labor supply, then the labor market becomes unstable. This is also mirrored by an awkward slope of goods supply--declining when price rises. The problem is partly solved when it is taken into consideration that labor income is constrained by superior bodies in a manner that takes only partial account of enterprise profits, because this drastically reduces the elasticity of labor demand and the likelihood that it may surpass that of supply. It is finally buried when we understand that a cooperative cannot reduce its membership when conditions improve, to benefit the remaining members. In other words, labor should be considered a fixed factor, like capital. One implication of course is that a labor-managed economy is less flexible than its capitalist equivalent. (15) Yet this problem underscores a real one that replaces it, that labor management leads to employment decisions that do not react to external signals. Thus there are no forces that drive the labor-managed market toward clearing and an equal-wage equilibrium, even where the LMF-market is fully employed, given the tastes of the LMFs that populate it.

Yet the distinct character of the Yugoslav system should not be lost: in the Soviet system additional labor is nearly costless, as far as the enterprise is concerned, and demand for it has no effective limit, eliminating perceived unemployment--unlike the Yugoslav-type labor-managed economy. Flence excess employment should be lower in Yugoslavia than in the other socialist countries.

The analysis explains the basis of certain well-known phenomena, stylized facts of the Yugoslav economic history. Thus persistent unemployment in most parts of the country, in spite of the extensive exports of labor to Western Europe, is based on the absence of any tendency to full-employment equilibrium in a Yugoslav-style market. Slovenia seems to be the exception, but the higher ratio of capital, human, and physical, to labor in this republic may have led to excess demand for labor and full employment. The same missing drive toward equilibrium, which is not mitigated by centralized wage setting, explains the wide and growing spread of incomes in different enterprises in similar industries and locations (Estrin, 1983 [2010]).

A closer look at the disintegration of the Yugoslav system exemplifies another aspect of the system. If workers believed that their place of work really belonged to them in spite of the legal fiction which declared all productive assets to be 'social', that is, belonging to society at large and not to the particular set of workers using them, then we would expect that privatization would be significantly harder than in other communist countries, where these assets belonged to the state. In particular, one would expect that the terms of privatization would provide worker-owners with a considerably higher share of privatized firms. The process of privatization in Yugoslavia is described in Uvalic (1997), (16) and it can be seen clearly that it was longer drawn out than in the other socialist countries and provided workers with considerably higher shares of ownership in the assets of the privatized entities.

And a last point: the Yugoslav experiment provides another instance of a general rule: serious limits on the operation of any of the markets, for either goods or primary resources, lead to inefficiency in the others. Socialist economies, be they of the fully centralized or partly decentralized type, impose strict constraints on the ownership of capital and thereby lead to centralization of firm entry and exit decisions. The capitalist economy's solution to unemployment of potentially productive workers in a regional market is to encourage the entry of profit-seeking entrepreneurs. It might be thought that in a labor managed market, groups of workers may obtain bank loans to start their own LMFs. Yet these workers' incentives are warped: since they cannot possess any capital of their own to endanger, their best bet is to disregard risks and go for the highest win, because the provider of capital, the community at large, will bear the loss if the bet goes sour (Keren and Levhari, 1992). Hence only a centralized capital allocation will survive (Keren, 1993).



Restate equation 2 as

[psi] [equivalent to] [Q.sub.L] - Q/L + FC/p/L [equivalent to] 0 (A.1)





[[psi].sub.p] = - FC/[p.sup.2]L < 0 (A.3)


dL/dp = - [[psi].sub.p]/[[psi].sub.L] < 0 (A.4)


Centralized wage setting

Consider equation 5, defining the setting of the administered y:

y = (1 - [alpha])[y.sub.-1]+[alpha]([pQ(L) - FC]/L) (B.1)

To maximize y, [y.sub.L] = 0:

[y.sub.L] = [alpha]([[pQ.sub.L]L - (pQ(L)-FC)]/[L.sup.2] = [alpha]/L ([pQ.sub.L] - y) = 0 (B.2)

and the optimizing L is defined by

y = (1 - [alpha])[y.sub.-1] + [alpha][pQ.sub.L] or [Q.sub.L] = 1/[alpha]p (y + [1 - [[alpha]]/[alpha]] [y.sub.-1]) (B.3)


This note has grown out of the writings of Milica Uvalic and Saul Estrin and is largely based on them. It has benefited from comments in a workshop at Aston University and by Milica Uvalic on an earlier version. I am very grateful to them and to three knowledgeable referees who helped correct several points, as well as to the editor, Paul Wachtel, who has taken great pains to ease the note's readability. None of them is of course responsible for any errors that might remain.


Berliner, JS. 1957: Factory and manager in the USSR. Harvard University Press: Cambridge, MA.

Berliner, JS. 1959: Managerial incentives and decision making: A comparison of the United States and the Soviet Union. In comparisons of the United States and Soviet economies, a compendium of papers submitted to the Joint Economic Committee of the Congress of the United States: Washington, pp. 349-376.

Byalsky, M, Keren, M and Levhari, D. 1999: The Kibbutz as a labor-managed club: Public and private goods, incentives and social control. In: Ranis, G and Raut, LK (eds). Trade, Growth and Development: Essays in Honor of Professor T.N. Srinivasan. Elsevier: Amsterdam.

Estrin, S. 1983[2010]: Self-management: Economic theory and Yugoslav practice. Cambridge University Press: Cambridge, UK.

Estrin, S and Uvalic, M. 2008: From Illyria towards capitalism: Did labour-management theory teach us anything about Yugoslavia and transition in its successor states? Comparative Economic Studies 50(4): 663-696.

Granick, D. 1972: Managerial comparisons of four developed countries. MIT Press: Cambridge, MA.

Horvat, B. 1972: Critical notes on the theory of the labour-managed firm and some macroeconomic implications. Economic Analysis 6(1): 3-4.

Keren, M. 1993: On the (im)possibility of market socialism. Eastern Economic Journal 19(3): 333-344.

Keren, M and Levhari, D. 1992: Some capital market failures in the socialist labor managed economy. Journal of Comparative Economics 16(4): 655-669.

Keren, M, Levhari, D and Byalsky, M. 2006: On the stability and viability of cooperatives: The Kibbutz as an example. Acta Oeconomica 56(3): 301-321.

Leppanena, S, Linden, M and Solanko, L. 2012: Firms, public good provision and institutional uncertainty: Evidence from Russia. Economic Systems 36(4): 522-530.

Levine, HS. 1959: The centralized planning of supply in Soviet industry. In comparisons of the United States and Soviet economies, a compendium of papers submitted to the Joint Economic Committee of the Congress of the United States: Washington, pp. 151-176.

Levine, HS. 1961: A Study in Economic Planning. Unpublished dissertation, Harvard University.

Meade, JE. 1972: The theory of labour-managed firms and of profit sharing. Economic Journal 82(325): 402-428.

Pencavel, J. 2012: Worker cooperatives and democratic governance, IZA Discussion Paper No. 6932, Bonn: Germany.

Portes, R. 1978a: Hungary: Economic performance, policy and prospects, in East European economies post-Helsinki. A compendium of papers submitted to the Joint Economic Committee of the Congress of the United States, Washington, pp. 766-815.

Portes, R. 1978b: Reply to professor Simai. ACES Bulletin 20(3-4): 81-84.

Prasnikar, J and Svejnar, J. 1988: Economic Behavior of Yugoslav Enterprises. In: Advances in the Economic Analysis of Participatory and Labor Managed Firms. Vol. 3, JAI Press: Greenwich, CT and London, pp. 237-311.

Robinson, J. 1967: The soviet collective farm as a producer cooperative: Comment. American Economic Review 57(1): 222-223.

Satt, E. 1996: Relative deprivation in the kibbutz economy: An exploration of the concepts of equality and equity. Economica 63(250, Supplement): S87-S101.

Shachmurove, Y and Spiegel, U. 1996: The optimum size of the labor-managed firm. Atlantic Economic Journal 24(2): 110-117.

Simai, M. 1978: Hungary in East European Economies Post-Helsinki, a Review Article. ACES Bulletin 20(3-4): 65-80.

Stiglitz, J and Weiss, A. 1981: Credit rationing in markets with imperfect information. American Economic Review 71(3): 393-410.

Uvalic, M. 1992: Investment and property rights in Yugoslavia. The long transition to a market economy. Cambridge University Press: Cambridge, UK.

Uvalic, M. 1997: Privatization in the Yugoslav Successor States: Converting Self-Management into Property Rights. In: Uvalic, M and Vaughan-Williams, D (eds). Privatization Surprises in Transition Economies: Employee-Ownership in Central and Eastern Europe. Edward Elgar: Cheltenham, UK.

Uvalic, M. 2010: Serbia's transition. Towards a better future. Palgrave Macmillan: London.

Uvalic, M and Vaughan-Whitehead, D. 1997: Privatization Surprises in Transition Economies: Employee-Ownership in Central and Eastern Europe. Edward Elgar: Cheltenham, UK.

Vanek, J. 1969: Decentralization under workers' management: A theoretical appraisal. American Economic Review 59(5): 1006-1014.

Vanek, J. 1970: The general theory of labor-managed market economies. Cornell University Press: Ithaca, NY.

Vodopivec, M. 1993: Determination of earnings in yugoslav firms: Can it be squared with labor management? Economic Development and Cultural Change 41 (3): 623-632.

Ward, BN. 1958: The firm in Illyria: Market syndicalism. American Economic Review 43(4): 566-589. Weitzman, ML. 1983: Some macroeconomic implications of alternative compensation systems. Economic Journal 93(372): 763-783.

Weitzman, ML. 1985: The simple macroeconomics of profit sharing. American Economic Review 75(5): 937-953.

Weitzman, M. 1990: Reflections on macroeconomics and share systems, Economic Journal Centenary. Basil Blackwell: Oxford, UK.

Woodward, SL. 1995: Socialist unemployment: The political economy of Yugoslavia, 1945-1990. Princeton University Press: Princeton, NJ.

(1) Their coauthored 2008 paper is based on their earlier research: on the labor market, by Estrin (1983 [2010]) and on the capital market by Uvalic's (1992). Although this note deals explicitly only with the labor market, it also embodies Uvalic's (1992) conclusions: the capital market in Yugoslavia did not conform with the conclusions of the models of Vanek, and of Pejovich and his collaborators, who based their analysis on the logic of Ward's LMF.

(2) Tyson reports on cases where delays occasioned by haggling organizations slowed down decision making.

(3) The fact of strong political influence was mentioned by most sources, starting with Ward (1958) who mentions the formal rights of intervention and ending with Prasnikar and Svejnar (1988) and Vodopivec (1993), who complain of excessive meddling.

(4) Leppanena et al. (2012) show that the symbiosis in the socialist countries between enterprises delivering public services, and local authorities offering protection, provided essential stability. This may be a factor that did not receive the attention it deserves.

(5) It was also used to analyze the Israeli kibbutz, (eg, Satt, 1996; Shachmurove and Spiegel, 1996; Byalsky et al., 1999; Keren et al., 2006).

(6) This section is based on Prasnikar and Svejnar (1988) as well as the writings of Estrin and Uvalic.

(7) An extensive historical outline of the Yugoslav system is provided by Estrin (1983 [2010], pp. 57-77) The few sentences below follow his account.

(8) Wording kindly suggeseted by a referee.

(9) This may have contributed to the recentralization in the 1970s, As Portes claims, a similar symptom led in 1972 to recentralization of the Hungarian New Economic Mechanism (see Portes 1978a, as well as Simai, 1978 and the reply in Portes, 1978b).

(10) Bankruptcies are analyzed in Uvalic (1992, pp. 106-122), especially Table 6.2 (p. 107). Their number increased at the end of the 1980s, but this can be seen as the start of transition from socialism in Yugoslavia (Uvalic, 1992, Chapter 9).

(11) Yet Horvat (1972), claims that the 'The Yugoslav experience shows ... chronic overemployment in the firms' (p. 290). I am not aware of any later reference to such a situation. Uvalid (private communication) explains that solidarity concerns may have led to some excess employment, that is, that one may expect to find that family members or friends are taken in even when not needed.

(12) Private communication from Milica Uvalic.

(13) Vodopivec (1993, p. 625) provides a table that relates the relation of the firm's personal income fund to its 'relative business success'.

(14) Slovenia had virtually full employment at a time when unemployment in the south of Yugoslavia was very high (see Table 1), possibly because of its higher endowment of capital, both human and physical.

(15) But cf. Weitzman (1983, 1985, 1990] who advocated using the share economy, a system very similar to the Wardian model, as an employment stabilizer in a capitalist economy. Note that the latter is not characterized by the specific organizational peculiarities of the Yugoslav type of economy that is the subject of this paper.

(16) And in Serbia in Chapters 6 and 7 in Uvalic (2010).


Economics, Hebrew University of Jerusalem, Mt. Scopus Campus, Jerusalem 91905, Israel.

Table 1: Unemployment by republic, 1959-1990

                            1959-1964   1965-1969   1970-1974

Yugoslavia                     6.3         9.1         7.4
Slovenia                       1.9         2.9         2.3
Other developed republics      5.4         6.5         6.5
Serbia proper                  6.2         8.6        10.3
Less developed republics       8.0        11.5        13.3
Kosovo                        20.3        28.4        22.8

                            1975-1979   1980-1984   1985-1990

Yugoslavia                    13.9        15.7        16.8
Slovenia                       1.6         1.7         2.6
Other developed republics     10.2        10.9        11.5
Serbia proper                 17.9        17.8        15.6
Less developed republics      20.2        22.2        24.1
Kosovo                        37.8        49.9        57.8

Source: Woodward (1995, Appendix Tables 6-9, 6-10 and 9-1); the
numbers in Table 1 are unweighted averages of the data in the
source. Legend: Other Developed Republics--Croatia and Vojvodina;
Less Developed republics--Bosnia-Herzegovina, Macedonia and
COPYRIGHT 2014 Association for Comparative Economic Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Keren, Michael
Publication:Comparative Economic Studies
Article Type:Report
Geographic Code:4E0EE
Date:Dec 1, 2014
Previous Article:The term structure of sovereign default risk in an emerging economy.
Next Article:Some observations on the convergence experience of Turkey.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters