Rationality and Distribution in the Socialist Economy
* PhD Thesis: Rationality and distribution in the socialist economy. Dapprich, Jan Philipp. University of Glasgow (2020).
URL = pdf download
Description
"The thesis provides a philosophically grounded account of a socialist planned economy. While I do not primarily consider a positive case for socialism, I address two major objections to it and thus argue that the possibility of socialism as an alternative form of economic organisation has been dismissed too quickly. Furthermore, I provide an account of the precise form a socialist economy should take, outlining general principles of planning and distribution.
Based on a welfarist interpretation of Marx, I show that distribution of consumer goods should be facilitated by an equal distribution of tokens. These tokens can be redeemed for consumer products or substituted for additional leisure time. The rates at which tokens can be redeemed for consumer products should correspond to market clearing prices. Welfare-oriented socialism is also defended against a deontological objection to socialism by Robert Nozick, who claims that socialism leads to injustice because it violates private property rights.
The thesis also considers Ludwig von Mises’s calculation argument against socialism, which claims that socialism leads to the abolition of economic rationality. I show how this objection can be overcome by using optimal planning techniques which are responsive to consumer demand as signaled by the market clearing rates of consumer products. The resulting model of socialism is tested using a computer simulation. The simulation also demonstrates that a novel system of valuation based on opportunity cost leads to a better adaptation of production in response to environmental constraints when compared to the labour values of classical political economy."
(https://theses.gla.ac.uk/81793/)
Contents
"In my account of a socialist planned economy, I intend to address arguments against such an economy, as well as questions regarding the precise form such an economy should take. In the case of the ‘calculation argument’ (von Mises 1920, see Chapter 6) against socialism, these two points overlap, since a successful objection to the argument mandates that socialism takes a form which does not fail on the grounds of economic rationality outlined by the argument.
This is not the case for the ‘deontological objection’ to socialist distribution by Nozick (1974, 162-164, see Chapter 3). This objection fails on its own grounds and my considerations of the distribution under socialism (see Chapters 4 and 5) thus do not pay much attention to it.
A consideration of an alternative form of social organisation, as presented in this thesis, can be called ‘utopian’ or ‘utopianism’. Utopianism has been criticised from various directions as being unscientific and detached from reality (cf. Nettl 1965, 73; Geoghegan 1987, 40; Popper 1945, 138-142; Popper 1957, 64-70). Chapter 2 therefore discusses these objections and argues that a scientific utopianism, as proposed by Neurath (2004a, 345-346), is possible. Scientific utopianism makes use of theoretical considerations and empirical foundations to design social institutions that best provide for social and individual needs. As with any scientific practice, scientific utopianism must include consideration of critical arguments. The so-called socialist calculation debate (von Mises 1920, Hayek 1981, Greenwood 2006, O’Neill 1996, Steele 1992, Neurath 2004b, Cockshott 2010) thus represents an exercise in scientific utopianism and such critical engagement should be encouraged. Proponents of a socialist society should address such arguments as I will do in Chapters 6 and 7. The purpose of Chapter 2 is to justify the project of developing a socialist utopia. Such a project is not doomed to fail from the beginning due to some systematic epistemic reason that might make rational engagement with alternative forms of social organisation impossible.
Chapters 3, 4 and 5 deal with the question of distributional justice. Chapter 3 provides a welfarist interpretation of Marx. According to this interpretation, economic systems based on exploitation should be rejected not because exploitation of workers constitutes a violation of workers’ entitlements, but because exploitation has negative consequences for their welfare. The purpose of the chapter is not to show that such an interpretation is the only possible or even the most literal interpretation of Marx. It is rather to show that a recognizably Marxist form of welfarism is possible and can provide a sensible basis for a discussion of distribution under socialism.
Chapter 3 also defends a welfare-oriented socialism against what I call the deontological objection to socialism, as formulated by Robert Nozick (1974, 162-164, see Section 3.3).
According to Nozick, it is not free market capitalism, but socialism that violates entitlements or property rights. Nozick thus suggests that socialism is necessarily unjust. I will argue that Nozick does not successfully demonstrate that such entitlements exist or should be taken seriously. We thus do not need to reject socialism on deontological grounds and distribution under socialism should not cater to the Nozickian concern for property rights.
In Chapters 4 and 5, I then consider the specific distributional mechanisms and patterns that ought to be implemented under socialism. The chapter assumes that consumer products are distributed to the population through tokens, similar to socialist labour token proposals (c.f. Marx 1999a, Part 1). This leaves open two questions which are adressed in Chapters 4 and 5 respectively. The first question is according to which patterns tokens should be distributed to the population. The second question is what token prices various consumer items should have, i.e. how many tokens need to be redeemed for the various kinds of items.
For the distribution of tokens, I argue for a version of Abba Lerner’s probabilistic egalitarianism (Lerner 1944, 25-30). Under ignorance about individual differences an equal distribution of tokens will lead to the best expected outcome in terms of welfare. While such a distribution should be adjusted according to available knowledge, the strict ignorance condition serves as a reasonable baseline for distribution. However, I suggest that individuals should be allowed to substitute tokens for additional leisure time. The consequence of this is that the tokens received by workers would be closely linked to their labour time. This is not argued due to some entitlement that labour creates but is based purely on the welfare benefits of such a distribution.
Token prices should be set at market clearing prices which balance supply and demand. This does not only lead to the most efficient distribution of consumer products, but also serves to conserve the egalitarian distribution of tokens. To argue this, I draw on Ronald Dworkin’s auction thought experiment (Dworkin 1981; Dworkin 2002, 65-73; Dworkin 2011, 356-358).
Dworkin recognised that a Walrasian auction (Walras 2014) is necessary in order to ensure that an equal distribution of tokens leads to an equal distribution of items. This is so because the market clearing prices resulting from such an auction are the most adequate measure of the opportunity cost that the appropriation of any one item inflicts upon others.
To summarise, in order to ensure the best welfare outcome, consumer products ought to be distributed equally before substitution between leisure and consumption has taken place. This can be achieved by giving each consumer an equal number of tokens. However, this is not enough to ensure equality. It is furthermore necessary to ensure that the token prices of items are set at market clearing rates.
Perhaps the most influential objection to socialism is Mises’ calculation argument (von Mises 1920, 1922), which commenced the wider socialist calculation debate (c.f. Hayek 1945, 1981, 2002, Lange 1936, Steele 1992, O’Neill 1996, Greenwood 2006, Cockshott 2010). Mises argues that in the absence of the monetary valuation provided by markets, including valuation of the factors of production, rational economic decision-making would be impossible. Chapter 6 discusses two interpretations of this argument. According to the first interpretation, Mises argument is about the nature of choice between competing ends. Mises claims that in order to make such a choice we must be able to commensurate the two options on a single scale. The question whether economic rationality is possible is about whether value judgements about ends are possible.
The alternative interpretation is that Mises’s argument is not about ends, but about instrumental rationality. Socialism fails not because it makes it impossible to choose between various ends, but because planners cannot determine the best means of realising these ends. In a complex industrial economy, there are countless possible uses for the various factors of production, such as machinery and labour. Without market clearing prices to guide them, planners would not be able to identify where these factors are needed the most and thus cannot make a rational decision about their employment.
In a sense, I am doing Mises a favour by accepting his premise that an economic system must allow for rational decision-making. An arguably much better way of evaluating systems is to consider their consequences. Socialism could in principle have superior consequences for many people’s welfare even if it does not allow for economic rationality. A poor person under capitalism, one might argue, is no better off because they can make rational decisions about how to spend their non-existent money. By accepting the framing of the debate between socialist planning and a capitalist market economy in terms of rationality, much of the misery of capitalism disappears. People under capitalism can seemingly make rational choices within the strict confinements presented to them. What those confinements are and what options people are left to choose from, however rationally, becomes irrelevant.
The reason I nonetheless accept the framing of the debate in terms of economic rationality is that socialists precisely claim (Engels 1962, 264) that socialist planning allows for a more conscious and rational configuration of the economy, a claim which I support. So, while the ultimate criterion by which to judge socialism should be its consequences for people’s welfare, I agree with Mises that the best outcome will likely be arrived at in a system in which productive factors are deployed in a rational manner. The question is thus whether Mises is correct in his claim that economic rationality is impossible under socialism. Chapter 6 argues that both interpretations of Mises’s argument fail individually, but that the strongest version of the calculation argument emerges when both interpretations are combined. The real challenge socialists must face is how to choose between competing ends, given the countless possible ways of achieving them, or between the various means, given the countless possible ends that could be achieved by them. One way of dealing with a vast number of possible plans is to use an optimisation algorithm, such as linear programming which directly solves for an optimised plan. The trouble is that this requires a prespecified objective, the fulfilment of which is to be optimised. I suggest that the choice of such a ‘plan target’ could be based on comparisons of market clearing prices and valuations that take into account the conditions of production. A similar proposal was made by Cockshott and Cottrell (Cockshott & Cottrell 1993). But while Cockshott and Cottrell rely on classical labour values, I propose a novel system of valuation based on opportunity cost. The resulting system chooses a plan under consideration of both the various possible ends and the available means and thereby offers a solution even to the strongest version of the calculation problem. An alternative version of the calculation argument, perhaps better described as the information argument, was later developed by Friedrich August von Hayek (Hayek 1945, 1981, 2002). Hayek emphasised the role that the market price mechanism has in conveying dispersed information. This information, according to Hayek, cannot be centralised for planning purposes. This is because the information is dispersed and because of its tacit nature, which supposedly makes it impossible to communicate it. Thus, socialist planners would necessarily lack the information they need to plan the economy.
I have not included a separate discussion of Hayek, because I consider the information argument to be much weaker than Mises’s calculation argument. Hayek remains rather mysterious about what kind of information he means, why planners would need it and why it cannot be communicated to them. I do not see why any of the information needed for the model developed here to work could not be aggregated at a planning agency or even made publicly available. A proponent of Hayek’s argument that wishes to criticise my model should point out precisely which information required in the model she thinks cannot be communicated, why it cannot be communicated and why a planning agency would need it in the first place. Technological circumstances have significantly changed since Mises and Hayek first questioned the feasibility of socialist planning. Modern computing and information technology can be used to transfer and process information much faster than ever before and this must be kept in mind when discussing the possibility of rational economic planning. I therefore suggest that the socialist economic system developed in this thesis be understood as an algorithmic process that could to a significant extent be run using computers and computer networks. This also makes it possible to simulate such a system on a computer, and a simplified simulation is outlined in Chapter 7. The idea that computers and computer networks could be used to control a socialist economy is almost as old as the computer itself. In the 1940s American mathematician and philosopher Norbert Wiener (1965) and his associates developed the field of cybernetics, which views machines, organisms and social structures as control systems involving complex flows of information. Experts in the Soviet Union quickly came to see this as the right approach to dealing with the increasing complexity of the Soviet economy, giving rise to the Soviet cybernetics movement (Gerovitch 2008, 2002). Long before the development of the internet, they put forward several never-to-be-realised proposals to build a nation-wide computer network that was to be used for both military and economic planning purposes. Actually put into practice was a much more low tech solution in Salvador Allende’s Chile (Medina 2011, 71-72). The project CyberSyn was developed by British Cybernetician Stafford Beer and connected telefax machines in various factories with a central control room in Chile’s presidential palace. From this control room the transportation of goods and materials could then be adjusted as new information came in.
Back in the Soviet Union, the idea of economic cybernetics was also thought in the context of Soviet research on optimal planning around the mathematician and economist Leonid Kantorovich (1960, 1965). Kantorovich first developed linear programming as an algorithmic method for solving optimisation problems he encountered in the Soviet economy (Kantorovich 1965, xvii-xix). While these problems could not be solved with conventional mathematical analysis, Kantorovich showed that they could be solved through a process of successive adjustments leading to an optimal solution. While linear programming can also be used to maximise profitability, Kantorovich used a non-monetary objective function. Linear programming can thus be applied in non-market settings, as it does not rely on market valuations.
The simplified model and its simulation outlined in Chapter 7 make use of linear programming with such a non-monetary objective function. Instead of monetary value, the physical output of a consumer good is maximised, while relative proportions of the output of various consumer products are fixed. The proportions are then successively adjusted through two feedback control loops which respond to the behaviour of consumers simulated by a simple agent-based model.
The purpose of the simulation is threefold. First of all, it demonstrates in more technical detail how the complete model developed throughout the thesis works. Secondly, it can be used to identify potential problems. And finally, I used it to specifically test how the model responds to environmental constraints and compared this to the labour value model of Cockshott & Cottrell (1993). The results clearly show that when a constraint on permissible greenhouse gas emissions is introduced, my model will often lead to a stronger shift in production towards low emission products. I argue that this shows that my opportunity cost valuations are a more adequate cost indicator than the labour values used in the Cockshott and Cottrell model. The overall structure of the thesis ties in together as follows. Chapter 2 serves to establish the project as legitimate. The rational and scientific study of alternative forms of social organisations is possible and sensible. Chapter 3 then looks at Marx’s views on distributive justice. While discussions about Marx and justice have mostly been framed around whether or not he thought capitalist exploitation was unjust, a welfarist interpretation is put forward which can also serve as a basis for distribution under socialism. Nozick’s deontological view on which he bases his objection to socialism is a stark contrast to this welfarist interpretation but is rejected after careful consideration. Chapters 4 and 5 then discusses precise principles for the distribution of consumer goods based on the general welfarist view of the previous chapter. Consumer goods are to be distributed based on an egalitarian distribution of tokens and clearing prices.
Chapter 6 then discusses the planning of production in the context of the calculation argument. The solution put forward draws on the clearing prices established in the previous chapter, as well as on optimal planning and opportunity cost valuation. This gives us the complete model which is then formulated in technical terms as part of the simulation presented in Chapter 7."
Excerpt
From the Introduction, by Jan Philipp Dapprich:
"The aim of this thesis is to provide a philosophically grounded account of a socialist planned economy. While I will not primarily consider a positive case for socialism, I will address two major objections to it and thus argue that the possibility of socialism as an alternative form of economic organisation has been dismissed too quickly. Furthermore, I will provide an account of the precise form a socialist economy should take, outlining general principles of planning and distribution.
The market as the predominant form of economic organisation has been a significant object of discussion and contention in western political philosophy (Herzog 2017). In the early Enlightenment, thinkers like Mandeville (1924) and Rousseau (1754) took up opposing views on the upcoming commercial society of their time. Marxists (Marx 2008, 192-330; Cohen 2014) criticise the exploitation they associate with capitalist market economies, while a variety of defences of the market have been offered (e.g. Nozick 1974, von Mises 1920, Hayek 1945, Friedman & Friedman 1962).
However, discussions of the market have not been limited to taking position for or against the market. Critical friends of the market (Herzog 2017, Section 2) acknowledge problems with market economies and discuss the extent to which other social institutions should interact with the market to counteract these problems. Discussion of market societies has thus not been limited to whether they are good or bad, but specific issues of justice within such market societies have received significant attention (e.g. Rawls 2001, Dworkin 2002).
In comparison, the philosophical literature dealing with socialist economic planning as an alternative to the market is not as rich and has not received as much attention. Such literature often deals with socialist planning in comparison to the market rather than in its own right. Planning is criticised because it supposedly does not have the same desirable qualities that markets have. For example, Ludwig von Mises and Friedrich von Hayek outlined how markets enable rational economic decision making (von Mises 1920, 1922) and convey economic information (Hayek 2002, 1945, 1981). The focus of their analysis is thus the market and socialist planning is only discussed as an antithesis to the market that supposedly does not share the same virtues. This predominant focus on the market has in my opinion contributed to socialism and socialist planning being dismissed too quickly.
There has been some literature on the question of distributive justice under socialism, most notably by Jerry Cohen (2009). Cohen advocates a kind of luck-egalitarianism, which disallows inequality based on people’s circumstances, rather than their choices (Gilabert & O’Neill 2019, Section 3.1). For other socialists, the starting point is Marx’s famous principle for what he calls the higher stage of communism: From Each According to his Abilities, to Each According to his Needs (Marx 1999a, Part 1; Carens 2003; Gilabert 2015). Less attention has been given to Marx’s description of the lower stage of communism (Marx 1999a, Part 1), which will be the starting point for my discussion of socialist distribution. Socialism has also been criticised as being in violation of principles of distributive justice by Robert Nozick (1974, 162-164).
Several models of socialism have been put forward (Gilabert & O’Neill 2019, Section 4), many of which compromise on the idea of (centralised) economic planning. Michael Albert’s Parecon (Albert 2003) for example emphasises the direct participation of citizens in the planning process. Under market socialist models central planning is to a significant extent replaced with market processes or trial and error processes mimicking markets (Dickinson 1930, 1939, Lange 1936, Miller 1981). These models largely accept Mises’s claim (von Mises 2008, 23) that markets are necessary for a rational economy. Some later models even allow for a form of private ownership of stocks (e.g. Roemer 1994). In the Soviet Union, some economists also emphasised the need for enterprises to operate to make a profit (Gloushkov 1969). The difference between capitalist profit and this socialist profit is that socialist profit is intended for projects benefiting workers. A model more clearly in the central planning tradition was put forward by Paul Cockshott and Allin Cottrell (Cockshott & Cottrell 1993) in their book Towards a New Socialism (TNS). The TNS model emphasises the potential of computer algorithms and information technology to aid in the planning of a socialist economy.
My thesis attempts to add to the existing literature by discussing issues of economic rationality and distribution in a planned socialist economy. Drawing on existing philosophical literature on distributional justice, I will propose principles of distribution for a socialist economy.
Furthermore, I will argue that the critique of socialism by von Mises (1920, 1922) should be rejected. Based on this discussion I will derive a model of a socialist economy which represents a modified version of the TNS model (Cockshott & Cottrell 1993). I shall focus on addressing objections to socialism and giving a positive account of reasonable principles of distribution and planning. I will thus not critically engage with alternative models of socialism, other than the TNS model (Cockshott & Cottrell 1993) which has had a significant influence on my model. While political philosophers will find much that they are familiar with, at times I will step onto territory that might more properly be described as political economy or simply economics. Such a multi-disciplinary approach is necessary to do justice to the underlying issues.
Defining Socialism
By socialism, I shall refer to a form of social organisation in which the means of production are publicly owned, production in individual enterprises and transfer of goods between them is to a significant extent directed by a universal plan for the whole economy, and there are no significant class distinctions between members of the society. Public ownership of the means of production implies that enterprises and their machinery, tools, resources and so on are collectively owned and controlled by all members of society or by social institutions accountable to them and operating on their behalf. Worker-ownership under which workers have a distinct ownership stake in the enterprises they work for (c.f. Ranis 2019) should be seen as distinct from truly public ownership and thus does not meet my definition of socialism. When all enterprises, including their assets and products, must be publicly owned, the transfer of products from a supplier of a product to a consuming enterprise implies no change in ownership rights and thus does not involve exchange or monetary payment. A socialist society as I understand it instead plans the production and transfer of such goods and instructs individual enterprises to act accordingly. Products are thus not bought or sold on an open market, but simply passed on from one enterprise to another. While all economies involve some form of planning of individual economic processes or enterprises, socialism extends this principle to the economy as a whole. This universal planning thus replaces the market exchange interactions between enterprises, which are typical of a capitalist economy.
Last but by no means least, a socialist society is a classless society. While people might still differ in terms of their roles in the economy – some people might do manual labour, while others hold managerial positions – there are no significant differences between these people in terms of rights, political power and access to economic goods. Unproductive members of society that live by the labour of others might do so for reasons of age or disability but not because they belong to some form of aristocracy given special privileges or because of interest on accumulated wealth.
All three criteria are necessary for a society to be socialist. Non-socialist societies might completely fulfil one or two of the criteria or partially fulfil all of them. But I do not consider them socialist unless all three criteria are fully met. For example, during wartime and during post-war reconstruction countries like the United States and France relied to a significant extent on planning to direct resources towards the war effort and reconstruction (de Carvalho 2019). However, neither did so by completely abolishing private ownership in the means of productions and both maintained significant class differences. The thesis does not contain discussion of historical or contemporary ‘socialist’ countries, so I will not comment on the extent to which these might or might not have fulfilled the three criteria."
(https://theses.gla.ac.uk/81793/8/2020DapprichPhD.pdf)
The Soviet Cybernetics Movement
Jan Philipp Dapprich:
"The idea that computers and computer networks could be used to control a socialist economy is almost as old as the computer itself. In the 1940s American mathematician and philosopher Norbert Wiener (1965) and his associates developed the field of cybernetics, which views machines, organisms and social structures as control systems involving complex flows of information. Experts in the Soviet Union quickly came to see this as the right approach to dealing with the increasing complexity of the Soviet economy, giving rise to the Soviet cybernetics movement (Gerovitch 2008, 2002). Long before the development of the internet, they put forward several never-to-be-realised proposals to build a nation-wide computer network that was to be used for both military and economic planning purposes. Actually put into practice was a much more low tech solution in Salvador Allende’s Chile (Medina 2011, 71-72). The project CyberSyn was developed by British Cybernetician Stafford Beer and connected telefax machines in various factories with a central control room in Chile’s presidential palace.
From this control room the transportation of goods and materials could then be adjusted as new information came in.
Back in the Soviet Union, the idea of economic cybernetics was also thought in the context of Soviet research on optimal planning around the mathematician and economist Leonid Kantorovich (1960, 1965). Kantorovich first developed linear programming as an algorithmic method for solving optimisation problems he encountered in the Soviet economy (Kantorovich 1965, xvii-xix). While these problems could not be solved with conventional mathematical analysis, Kantorovich showed that they could be solved through a process of successive adjustments leading to an optimal solution. While linear programming can also be used to maximise profitability, Kantorovich used a non-monetary objective function. Linear programming can thus be applied in non-market settings, as it does not rely on market valuations.
(https://theses.gla.ac.uk/81793/8/2020DapprichPhD.pdf)
Discussion
A critique of Dapprich’s model
Simon Sutterlütti:
"Dapprich’s doctor father is Paul Cockshott, the well-known computer socialist and co-author of ‘Towards a new socialism’, and his work is largely in line with Cockshott’s ideas. Cockshott’s general critique of real socialism include at least three points: anti-democratic, bad planning (balanced plan instead of optimized plan, money-based instead of work-time-based – well, and now we have computers), no market for consumer goods (Cockshott & co want the prices for consumer goods be fixed by supply and demand, this information should be used to evaluate the different goods and do better planning and distribution of the means of production). The workers get tokens for the work they perform and with these they can buy goods. Tokens are not money, because they don’t circulate. Which (I think) essentially means you can’t buy means of production or workers with them. They are more like theater tickets, the theater can’t use the ticket to pay the workers or buy stuff.
Dapprich grounds his socialism in a utilitarian theory of justice and welfare. With Abba Lerner he argues that the best way to maximize welfare would be an equal distribution of income and leisure. Lerner assumes the diminishing marginal utility of income: the more income you have got, the lesser will be the satisfaction of an increase of income. “An individual will first use their income to satisfy their most urgent needs and then spend any additional income on less urgent ones“ (Dapprich 2020: 53). Therefore, equality of income maximizes welfare. In a second step, he argues that workers should be allowed to substitute income for leisure. And with leisure he also expects a decreasing marginal utility: “An individual will spend the first hours of leisure time on the activities most enjoyable to her” (cf. 57)[1]. Allowing the substitution between leisure time and additional income leads to a scenario “where individual income is proportional to the number of hours worked“(cf. 58) – the famous socialist principle “to each according to her contribution”.
Quite interestingly, Dapprich doesn’t arrive at this principle because he wants or needs income as an incentive to work, but from an entirely welfare-based argument. “But such a principle is not justified through rights or entitlements supposedly gained by labour, but by the welfare that the implementation of such a principle would lead to“ (cf. 70). He even mentions incentives as something “that somehow violates the principle of equality” (cf. 59).[2] In fact, in the discussion at the Future Histories Podcast Dapprich is asked the very same question we get all the time: Who does the garbage collection and other undesired jobs? Likewise Cockshott/Cottrell and Michael Albert (Parecon) argue – based on an idea of justice – that all work should be equally rewarded and only (small) differentiations due to performance should be allowed. Dapprich argues that a differential pay may be appropriate for unpopular jobs. Finally, the model includes common funds, that finance expansion of production, administration, basic services such as health care, etc."
(https://keimform.de/2021/critic-of-wage-based-cyber-socialism/)