Socialism Without Markets: Democratic Planned Socialism

Al Campbel, 2003

Note: As of 3/15, the paper below only has the introductory material and the first major part of the paper, the positive statement of the economic aspects of my model. The second major part, the discussion of three other models of DPS, will be [is planned to be] available shortly before the conference.


1. Two interrelated issues inform the debate between socialists who advocate Market Socialism and those who advocate socialism without markets, Democratic Planned Socialism (DPS). [1] The first is if socialism without markets is feasible. David Schweickart, who calls his most left wing of all market socialist models "Economic Democracy," flatly asserts: market socialism "is the only form of socialism that is, at the present stage of human development, ...viable..." (1998:10). From a logical point of view, one should as well consider the question of if market socialism is possible. In fact, a few advocates of socialism without markets have made that point, for example Bertell Ollman (1988) and David McNally (1993). By-in-large, however, this issue has not been included on the debate menu, while the question of the feasibility of socialism without markets has: advocates of socialism without markets have felt it necessary to defend their vision as feasible, while advocates of market socialism have not felt the same necessity. The second issue is, if both models are in fact possible, which would be more desirable.

2. Early attacks on capitalism, precursors of the modern socialist movement, were almost all strictly anti-market, based on the dehumanizing effects of markets, and distribution was to be affected according to need. [2] Similarly, the early socialists identified (generalized) markets with capitalism, and hence were clearly anti market. The subject of this conference, Marx, was very clear in his opposition to markets in socialism (1875). [3] Other early socialists who outlined something of their vision of a socialist society without markets were Bellamy (1888), Kautsky (1892), Bukharin and Preobrazhensky (1918), and Neurath (1919).

3. These early visions of socialist societies without markets were attacked by opponents for being very vague on a number of central issues concerning how they would function: how the non market planning would be done, how people would be remunerated, how various different actors would be motivated to do what the plans were based on them doing, and so on. Recently, a number of people have put forward extensively worked out models that address these issues at a mid level of abstraction, much more concretely than the earlier generation of non market socialist visions.

4. This paper is not fundamentally about the debate between market socialists and socialists who reject markets. [4] It is the position of this paper that there not only is a way to organize a socialist economy and society without markets, there are many ways it could be done. This paper will discuss four of these. In doing so, this will in passing underline the position in the debate with the market socialists that, contrary to the assertion of Schweickart, there indeed is a way (in fact many ways) to organize a feasible non market socialist economy. But for the interests of the author, that is considered a given, and the purpose of this paper is rather to consider the relative desirability of various aspects of anon market socialist economy, aspects which the models considered handle differently. [5]

5. While there are other recent works advocating various visions of non market socialism, [6] the three that have been outline the most fully in the published literature are those by Devine (1988), Albert and Hahnel (1991a and 1991b), Cockshott and Cottrell (1993).

6. These three models and mine that I will consider share more than being models of socialism without markets. I consider all of them to be DPS models. Since I will be arguing that some aspects of some of the models are less desirable than other ways of achieving the same goals, I want to stress two things in passing to try to assure that the reader is not left with an incorrect impression of my view of these models: 1) all of these models are more desirable than either capitalism or any version of market socialism, and 2) all of these models are broadly viable.

7. This paper will have two main parts. Following a brief but necessary discussion of the goals of socialism, the first main part will be an outline of economic procedures that I maintain would yield a DPS that would be both feasible and desirable. In the second main part I will discuss some aspects of the three models worked out above that I think are not as desirable as certain alternatives. I will not try to give a full discussion of these models. On the one hand, to do so would require a book length work. But beyond that, I not only agree with the general thrust of all three of these models, I agree with many of their specific proposals, either as the best way to achieve something, or more often as at least a good way among a number of possible good ways to achieve something. Hence my discussion will be focused very largely on what I disagree with, which are only small, though sometimes important, parts of the models..


8. At the broadest and most abstract level, the central goal of socialism has always been something like "human development," "the development of one’s human potential" or "the opportunity to develop potential abilities." At a slightly less abstract level, self governance (or often simply "democracy"), equality, and solidarity are the most commonly cited sub goals. Other still more concrete goals were intended to contribute to these goals, such as the standard (until recent Market Socialist times) goal of nationalizing the means of production, which was intended to contribute to both equality and self governance in the economy. Various authors list other goals they ascribe to socialism, such as "individuality" and "privacy" (Weisskopf, 1992a, 1992b), "liberty" and "autonomy" (Schweickart, 1996), and "variety" (Albert and Hahnel, 1991a, 1991b), but the traditional ones are still the ones most often referred to. Recently most socialist models, including the models of Democratic Planned Socialism referred to above, have included protecting the natural environment as an important goal.


9. Here I will discuss sixteen specific proposals for rules and procedures for democratic coordination, control and planning of the economy. There are other aspects to be considered in a full model, but these will be sufficient to convey the nature of the proposed model. The economy in DPS will differ from markets in the manner of determination of three central aspects: what is produced, how necessary inputs and human labor are brought to the production process, and how what is produced is distributed.

10. Under capitalism, what is produced is determined by profit maximizing companies. They respond to whatever direct or derived demand they believe they can make the most profit from, and they are subject to the laws of the market that form the environment in which they make their decisions on production. There is no pretense of democratic control of the economy.

11. Two basic types of democratic changes must be effected to establish popular control over the whole economy. On the one hand, decisions by the enterprises and organizations that produce society’s desired goods and services must become democratic, being made by those strongly affected by the decisions: certainly the workers in the enterprise, but in many cases also other larger bodies, such as the people that live near the production site. On the other hand, society’s members must establish democratic control over the interaction and coordination of these enterprise level decisions, and control over the aggregate results of these myriad enterprise decisions, to complete the popular democratic control of the whole economy. The failure to be concerned with this latter necessity is the fundamental weakness of market socialism. Two different types of mechanisms will together generate this social control: the direct democratic determination of a few socially important aggregates, and the specification of certain parametric algorithms for a number of enterprise decisions. This section will elaborate on all of these points.


12. People hold different opinions concerning what part of total yearly production should go to "the present generation," that is, consumption, and what part should be used to create a better economy for "future generations," that is, investment. Similarly, people hold different opinions concerning the ratios they would like to see between the three components of present consumption; individual consumption (consumer goods and services), collective consumption (for example national and local parks and other recreational facilities), and social services (education and health care would be two major ones). Therefore,

13. Procedure #1: The national population will vote to directly determine how to divide current GDP between present consumption and investment, and how to divide current consumption between individual consumption, collective consumption and social services and government operating costs. [7]

14. Of course, some procedure would have to be developed to enforce and enact these and all the other democratic decisions in society. This issue is important, but it is a general issue in democratic theory (in theory of concern even under liberal capitalism, for the political sphere), and will not be discussed in this work.

15. Procedure #2: Workers will be paid (collectively) the full value of what they produce (wages to be discussed below), and then taxed in accord with their vote just discussed. [8]

16. For example, suppose people voted for 10% investment and 90% consumption, and they voted for the division among current consumption to be 30% for social services, 15% for collective consumption, 5% for government operating costs, and for 50% individual consumption. Then taxes would take a total of 55% of GDP which would be spent according to (as a percent of total GDP) 10% for investment, [9] 27% for social services, 13.5% for collective consumption and 4.5% for operating the government, leaving 45% to be spent individually on consumer goods and services. Note in particular that this would ensure that the amount of money in the economy available for purchasing consumer goods and services would just equal the value of those goods to be purchased, so there would be no reason for demand pull inflation [10] and the devaluation of the "money." [11]


17. Traditional socialist models have differed on where a number of microeconomic decisions should be made. In particular, there have been differences concerning if some decisions should be made in an enterprise (by workers councils) or if they should be made at a supra enterprise level by planners who represent [12] a larger constituency or perhaps the whole nation. The general criteria for deciding this issue are

18. Procedure #3:

Decisions whose effects are predominantly internal to the workplace will be made by the workplace worker’s councils. [13]


19. Procedure #4: Sections of society larger than the workplace workforce must be included in any production decisions which significantly affect these larger sections of society. Such decisions, including the obviously potentially contentious one of what groups have sufficient stakes to be assigned representation (and how much representation), will be made by a democratically elected government or by boards democratically elected to carry out the task of making these choices to best reflect society’s members’ preferences.

20. It is important to understand the tremendous amount of additional self governance this would bring into people’s lives. The following two groups of types of decisions, presently all determined by owners of capital, would be governed by Procedure #3 and determined by workplace collectives.

21. Group 1. Relationship of workers to their workplace collective: hiring and firing; discipline; promotions, evaluation and training; transfers and leaves; internal information and communication systems; administrative procedures and rules; organizational form; extent and nature of supervision.

22. Group 2. Relationship of workers to one another and to the physical features of work: quality control; working conditions; methods of remuneration; maintenance of machinery and equipment; work methods, task ordering, job division, job rotation, variety of tasks, and so on; scheduling; work distribution and assignments; type and level of interaction among workers; employment of technology (that does not seriously impact the physical environment); non monetary incentives.


23. Other decisions would directly impact larger segments of the population. As an example, consider the adoption of a technology that might pollute the surroundingneighborhood, or might significantly contribute to national or global pollution. Here the extension of self governance to those significantly affected requires decisions be made at a supra enterprise level, as proposed by Procedure #4. Additionally, the collective consumption decisions and decisions on the amount and nature of social services to be provided discussed above should be made the same way.

24. Procedure #5:

Choices concerning investment, collective consumption and social services will be determined by a democratically elected government or by boards democratically elected to carry out the task of making these choices to best reflect society’s members’ preferences.

25. Comment #1.

One ongoing discussion concerning socialist economic models concerns centralization of decision making v.s. decentralization. This model clearly contains both centralized and decentralized decision locations. The important issue is what the criteria are for deciding how centralized/decentralized a decision will be.

26. The need for coordination is the main reason for requiring some level of centralization. If everybody buys a car because with the existing roads they can get to work faster than with a bus, the roads will end up being choked and the people will not get the rapid transportation that they chose to buy the car to achieve. Decentralization there does not do a good job of satisfying people’s preferences. If 20 steel producing plants across the country in a Market Socialist system see steel is selling well above cost and hence decide to invest to double their capacity to reap large profits on the invested capital, the market will be flooded, steel will no longer sell above cost, the investor collectives will not realize the goal they invested for, and society will have wasted resources.

27. Beyond the issue of collective self governance by people of the institutions they are part of, the main reason for decentralization is access to necessary detailed information. If one looks at the list of production decisions above, one can see that the workers in the enterprises themselves are the people who will have the knowledge required for many of the decisions. One could have this information relayed to a center, as was done for many of these decisions in the Bureaucratically Planned economy of the USSR. But depending on what incentive systems one had for the people involved, one could have deliberately incorrect information relayed to the center, as was in fact a major problem in the USSR, greatly diminishing the value of decisions made by the center.

28. The location of decisions on the centralize/decentralize spectrum should be determined by the economic nature of the decision being considered. In particular, decisions that require extensive coordination to achieve a socially optimal outcome must be sufficiently centralized, while decisions that need extensive and detailed local information and do not have severe coordination issues must be sufficiently decentralized.

29. Comment #2.

Most authors who write about a post capitalist non market socialist economy have stressed the importance to authentic human development of a profound transformation of the nature of work. The control given to workers’ councils in Procedure #3 above implies this deep change. There is not space here to elaborate on this, but it is important to emphasize its centrality to a socialist transformation. All DPS models refer to this, but it is addressed at greatest length in the works by Devine (1988) and Albert and Hahnel (1992a).

30. I want to next deal specifically with four decisions key to any economy involving a division of labor and exchange: wages, prices, investment and output. Note that under capitalism all are determined by (conceptually) simple algorithms, which all aim to serve the goal of maximizing enterprise profits. Algorithms for these four quantities will play an important role, though they certainly are not the only contributing factors as we have already seen, to the economic coordination of this model of DPS.

31. Goods and services produced will have exchange prices attached to them, and as the name exchange price suggests, the ratios of these prices will determine the amount of a good exchanged for another good or exchanged for money. Exchange prices certainly will want to be set to (roughly) balance the supply and demand for goods: shortages or surpluses represent wasted human time and wasted resources that could have been used to further human development. But the requirement that supply equal demand at a given price does not close the problem mathematically. For example, if one were at one price and had supply equal demand, and then producers decided they wanted to supply more output at every potential price, then one would move to a new, lower, price at which supply equaled demand. Having supply equal demand does not by itself determine the price.

32. The socialist goal of equality suggests that if a person contributes a certain number of hours to social production, she should be able to get in return goods and services that took the same amount of hours of labor by other humans to produce. In this sense, everyone’s time is held to be of equal value. [15] Together the wage, price and investment procedures presented below will achieve this egalitarian treatment of human labor.

33. Procedure #6. Every person will be paid the same amount per hour contributed of social labor.

34. While it clearly makes no difference mathematically if we call the wage $15/ hour or one labor credit/ hour, I advocate a wage rate of one labor credit per hour for reasons of transparency and consciousness.

35. Every good will carry two prices, an exchange price at which it will actually exchange, and a cost price.

36. Procedure #7. The exchange price for each good is set by each firm that produces it to clear the market for its good, as described below in discussing the decision by each firm on its output.

37. Procedure #8. The cost price of a good (consumer good, capital good, or intermediate good) will be the sum of the wages paid to the workers to produce it, the cost price of intermediate goods used, and the cost for the use of capital goods (see below on investment for this cost).

38. A major difference between this system and a market system enters at this point, and one that is particularly important to the ecological destruction that is occurring today. As has been repeatedly observed by its critics, neoclassical economics "high theory" largely ignores externalities. [16] For example, a production process can pollute, seriously harming the health of millions of people. The laws of the market prevent the company from spending money to return the environment to its original state even if it were inclined to do so, since that would raise its price and cut into its market share and profits. In the DPS system described here, the solution would be to simply require the enterprise to correct any damages to the environment from its production process and include the costs of doing so as part of the cost structure associated with that technology. [17] Note that this and most externalities affect many more people than the workers in the workplace, so the amount of environmental protection required would be another issue that would have to be determined above the enterprise level. Democratically selected experts or the affected population itself would determine the level of pollution that they considered non damaging to the environment.

39. With the exchange price and the cost price determined, we can now describe the investment process. Actually, there are three components to investment, and the exchange price/ cost price duality only is involved in one, the adjustments of the relative weight of production between branches in response to changing consumer preferences. The nature of the investment process is key to the viability of any non market (or market, for that matter) economy.

40. Procedure #9. A democratically determined "Investment Council" (IC) will determine investment. Recall that society has democratically determined the fraction of GDP it chooses to allocate to investment. The IC will first have to divide that amount among three distinct investment uses, and then decide how to further allocate the amount allocated to each use to specific projects.

41. The IC must first decide how to divide the investment funds among research and development (R&D), replacing existing capital with labor saving capital, and investing and divesting in various sectors to adjust output in accord with people’s constantly changing preferences. Note that while modelers sometimes concentrate on the latter, and that is important, that focus to the exclusion of the others is a reflection of the neoclassical attention to "static optimal allocation" concerns at the expense of real dynamic growth and new technology concerns. It is the former two that have the most importance for long term improvement of the efficiency of human labor, a key consideration (not the only) in possible social transformations.

42. #9a.

Two considerations must go into the decision about what part of the total investment funds to allocate to R&D. The first is a technical issue. How much return will one get for a given amount of R&D? This is a very complicated issue, because there is not even a single output from R&D that one is interested in. R&D could be engaged in for at least three different reasons: to raise labor efficiency, to mechanize types of work that are considered dangerous, unhealthy, or simply dehumanizing, or to protect the environment. The second consideration is that, providing one can get more of one or all three of these for more R&D spending, that still leaves open the issue of how much one cares about those gains compared to gains from spending on other things. How important is it to raise the growth of productivity from 2% per year to 4% per year, compared to the gains from shifting production to closer match people’s desire for a different pattern of current consumption by investing to change the productive capacities in different branches of the economy? That of course is a social value judgement.

43. #9b

. In deciding how much of the investment funds to allocate to replacing existing capital with more labor efficient capital, one of the two issues above does not arise. Presumably one can calculate how much human labor would be saved by a certain investment. That leaves the IC, however, still facing the social choice: what is the appropriate trade off between saving labor and restructuring the productive capacity more closely to the current consumption demand?

44. #9c.

As indicated in the discussions of points #9a and #9b, the IC needs to make a social value judgement on how to divide the investment funds between R&D, investment in labor saving technologies, and investment to realign the reproductive capacity more optimally with the current consumption demand. Once that social decision has been made, however, the following algorithm will optimally determine which branches of production get how much of those funds allocated for altering the relative productive capacities. Investment priority between branches will be according to the amount of price reduction in the exchange price times the amount of people who will benefit from this, [18] for a given investment expenditure. This is socially optimal because of the economic meaning of the algorithm. The exchange price indicates how much of the labor time which they have contributed to social production people are willing to pay for a good, or roughly, how long they would be willing to work to make it. The cost price reflects how much social labor it actually takes to make. The bigger the relative gap between these for a given good, the greater the gains to society from producing more of that good.

45. Note that all capital goods will belong to the people collectively, not to the workplace that uses them.

46. Procedure #10. A cost price will be calculated for capital goods just as for other goods. Recall every year the IC receives some democratically determined part of the GDP for investment. Once it has decided how to allocate that as described in the last procedure, it will purchase capital goods from capital goods producers, at cost prices, and distribute these to workplaces. Once a capital good is given to a workplace to use, a rent will be charged. That rent will be set to pay back the cost of the capital good to the IC, over the time they estimate it will take to be completely depreciated (from physical wear and tear, or from obsolescence). [19]

47. An important caveat is needed concerning the price mechanism just described, a second procedure motivated by concern with the rapidly growing environment crisis. If the cost of a limited harvest good (such as fish or timber) was such that at that price the demand was more than could be sustained over time, the resource would be depleted. Aside from its economic impact, that could be considered environmentally unacceptable.

48. Procedure #11. A tariff will be added to both the exchange price and the cost price of any renewable resource threatened with over harvesting to raise the exchange price to a level such that demand at that price will not exceed a level of production that is environmentally sustainable. For non renewable resources the same procedure would be used, where the level of production is set to a socially determined acceptable rate of depletion, including possibly a rate of zero if so desired.

49. This of course will produce a revenue for the government. That revenue could be used, for example, to lessen the tax needed to run the government. The use of the revenue, however, is a strictly secondary consideration. The system of incomes and expenses is already balanced without this revenue, and the point of the tariffs is to protect the environment.

50. Finally, consider enterprise output determination. As long as the exchange price is above the cost price, people in society are indicating that they would be willing to contribute more hours of their time to social production than it actually takes society to produce the good.

51. In general, an enterprise will face some downward sloping demand curve for its product, and it will have an upward sloping exchange cost curve. [20]

52. Procedure #12. Enterprises set their output and their exchange price so that their price equals their marginal exchange cost at that quantity of output. [21]

53. Note in passing that to the extent that large amounts of capitalist production takes place in oligopolistic industries, this DPS procedure would 1) yield important social efficiency gains over capitalism (and over Market Socialism), [22] and 2) represent a more authentic "consumer sovereignty" (relative amounts of consumer goods produced match consumers’ willingness to exchange their labor for them) than capitalism.

54. Just as models of DPS recognize that people differ in their ranking of social goods and services for their consumption, such models should recognize that people differ in their preferences concerning how much work and what type of work they desire to engage in. I want to end this discussion of procedures by very briefly indicating four procedures that would increase people’s choices concerning how they worked.

55. Procedure #13. Labor/Leisure tradeoff. People can work as many or as few hours as they choose in social labor.

56. This is important to best meet the spectrum of desires that people have on their labor/leisure (or even social labor/ "individual labor") tradeoff. Leaving aside the issue of access to free goods such as education and health care that society would have to decide on for people who chose to do minimal or no social labor, people who choose to work less still only draw back from the social collective what they contribute. As such they do not constitute an economic problem. Note that the labor/leisure tradeoff is simultaneously a high/low social goods consumption tradeoff, again something about which people will have different preferences.

57. Procedure #14. "Undesirable work" would earn some number of labor credits greater than one per hour, with the rate set to assure that the number of people desiring to do a certain type of work matched the number needed by society for the socially desired social product.

58. This of course runs counter to the egalitarian Procedure #6. It is seen as something only affecting a relatively small number of particularly undesirable jobs, as the only way for these few jobs to avoid conscripting labor, which I view as more socially harmful than the non egalitarian consequences of this procedure. However, the egalitarian Procedure # 6 remains the goal, and to constantly try to move toward it one has

59. Procedure #15. The greater the labor multiplier for some undesirable job, the more research efforts and funds would be directed toward restructuring or transforming the work to make it less undesirable, or mechanizing it to eliminate it.

60. This would tend over time to move any non unitary multiplier toward the standard value of one.

61. The final procedure concerns a different aspect of labor.

62. Procedure #16. Pay for Childcare. Any socially useful service, as determined by society, that does not produce a service that is paid for, should nevertheless be paid by society in accord with the logic of pay for social contribution. This is already envisioned by most advocates of socialism for free healthcare and education. The same logic should be extended to child rearing-people engaged in that should receive pay for their labor from society.

63. There are of course many issues here. Determining the rate of pay for such work in the home would require social discussion, since one is doing childcare all night long when one sleeps, but the nature of the work is quite different from most other social work. Further, the nature of raising children and the nature of allocating adult human time to that activity will certainly change radically under any socialist system from childcare work as it exists today. Again, all those are (complicated) details to be dealt with by the people involved. The point here is that raising the next generation is clearly a completely necessary social activity, and so it should be treated and recognized as such, and a non market system lends itself to doing that in a way that markets do not.