In his critical review of ‘Economic Planning in the Age of Climate Crisis,’ Renato Flores is left with more questions than satisfying answers from Cockshott et al.
Jean Tinguely, Meta-Chaos, 1972
Paul Cockshott and Allin Cottrell’s Towards a New Socialism has been circulating among the communists since 1993. It is a valuable book; it revitalised the planning debate by offering a series of simple solutions and algorithms. People could try the algorithms on their own computers. They showed through maths why a socialist society could mean less working hours for all. The core of TaNS’s proposal for a “new socialism” is to plan society using work hours as a basic accounting unit. In most instances, people would be remunerated in labour vouchers according to how much they work. With a series of restrictions on labour vouchers, they can avoid becoming a universal equivalent, i.e. just another type of money. The book’s political program is not limited to its proposals for managing resources: among other things, TaNS proposes the communalization of care work and the use of sortition as a substitute for political representation. The authors leave interesting threads which continue to inspire debate in the present.
Cockshott and Cotrell recently joined efforts with Jan Philipp Dapprich, another scholar in the planning tradition, to produce a new book on planning that incorporates ecological themes into their framework. Bracketing Cockshott’s more controversial (and, at times, frankly reactionary) contributions to other areas, I was excited to read this new book. TaNS is an inspiring read to which I occasionally return, despite not agreeing with every point. Plus, in a world where ecological collapse seems the most pressing challenge society faces, contributions which raise the level of debate are always welcome. But, after an initial skim, I worried Economic Planning in an Age of Climate Crisis fell rather short of its predecessor. And after a more in-depth read, my initial fears were confirmed: even if the (incremental) developments of the planning algorithms included in the book are interesting, there are no new serious proposals inside. It is a marginal contribution to finding solutions for the present ecological crisis, but in some respects it’s a downgrade from TaNS.
The book has nine chapters and one appendix with computer code. The first chapters are literature reviews of climate science where I have little to object, even if the authors underestimate the breadth of the ecological crisis by reducing it to a climate/carbon crisis, something already apparent in the title. This is a weakness of the book which is not unique to the authors: in too many recent eco-socialist books, the problem of ecological collapse is reduced to one of climate and carbon emissions. Climate is surely a large component of the ecological crisis, but it is not the entirety of the problem. This flaw could, in principle, be worked around – after all, the authors do talk about planning algorithms that avoid over-exploitation of non-renewable resources. One could add here land nutrients or use of certain rare metals as variables in the model. However, the problem is deeper than just accounting with more variables – it is a problem of method. And this becomes clear as the chapters pass: the authors clearly fail to outline a positive and actionable ecological program that goes beyond constraints in linear programming models. Only in the last four pages do the authors give some (rather weak) suggestions, ones which end up reproducing the logic of neo-liberal ecological economic methods like carbon pricing that assign a “value” to pollution.
One could argue that the book simply proposes planning algorithms, or a defense of planning, for the age of climate crisis, and it is up to us to pick up the threads and tie it all into a nice whole. However, the number of loose ends is not small, and tying them up is a work of very high order. This limitation was already present in TaNS, which has led to many misinterpretations of the book as a call for an “algo-cratic” central planner (that is, political life being subordinated to algorithms), or mistaking the entire system for market socialism behind a veil. The failure to address these gaps makes reading the new book even more frustrating: thirty years have passed since TaNS and the few algorithms the authors offer are not a significant step that takes us closer to a New Socialism. Instead, with the additional ecological dimension, the task has grown and is now even more formidable. And by removing the valuable discussions on democracy present in TaNS, the amount of practical tools the authors give us became practically zero.
The gap between Economic Planning and a workable ecological program is indicative of deeper issues. What is at stake here is the abyss between a more “math”-oriented branch of cybernetic planning and a more “systems”-oriented branch of cybernetic planning, and this is something I hope to bring to light in this review. The beauty of C&C’s approach is that it provides a method for deflating the calculation problem and gives us quantitative variables to work with. But it lacks serious discussions of implementation at different levels of organization, and provides little flexibility for multivariate accounting systems, ones which could work better at smaller scales and are more consonant with the environment. This is something the “systems”-oriented planning methods do much better: they focus on supervising the existing self-organized systems rather than managing society. However, these methods often fall into localism without providing a viable answer on how to organise the world. And when they do, it usually takes the shape of recursive models, such as attempts to concretize Stafford Beer’s Viable System Model. While Beer himself opposed prescriptive formulas for concrete organization and combined tools from different fields, some of his followers transform his work into an abstract schema which does not address the necessity of different approaches at different scales nor what types of information can be transmitted to avoid dilution and address the calculation problem.1
Between these two worlds lies this review of Economic Planning. Needless to say, I will not solve the issues at stake. However, I do hope to make them clearer so others can take their chances at contributing to a solution. The planning community is seeing very active debate, but I think it often focuses on the wrong things in a way that misses the forest for the trees. Here is my attempt to describe the forest.
Planning vs The Calculation Problem
My review starts on page 66, with the section “Countering the Austrian Critique”. Here, the authors recapitulate the “Austrian” arguments against planning, also known as the calculation problem or debate. As posed by Ludwig von Mises, the original economic calculation problem states that central planning can never be a substitute for the market, as planners lack an objective measure of value and thus cannot rationally allocate between producer and capital goods or effectively discipline producers. A further development of this problem is Friedrich Hayek’s formulation, which is that a planned economy misses the role of the entrepreneur and can never go through the innovation or creative destruction characteristic of capitalism because this role is absent. Furthermore, Hayek critiques that planners have an information problem, and cannot know the real preferences of consumers. The book addresses the Austrian formulation of the planning program at the end of Ch. 3 (pp. 66-71), with arguments by the authors available elsewhere in more detail. Von Mises is dealt with by stating that labor-time can be taken as this objective method for value. On the other hand, to Hayek’s objection, the authors counterpose the abundant historical experience of state-driven innovation, especially in times of crisis, and postulate that there are other methods for gathering information from consumers aside from the act of trading a commodity for money.
In Chapter 4, “Wartime Planning”, the authors show optimism for applying exceptional planning efforts to the current existential climate crisis. And in Chapter 5, “Post-War Experience with Planning”, the authors give further examples of successful planning after the Second World War; they show that their schemes are not utopian but have a basis in history, and that planning efforts are still widely applied today. I have little to object here: I believe that planning will be necessary to overcome the climate crisis; I also agree that a lot of scientific innovation happens through public means, and that most businesses incorporate heavy planning within them.
The book follows with Chapter 6, “Theory of Optimal Planning”, where the authors return to the area they are strongest at: they describe algorithms for planning increasingly complex problems, which were not present in TaNS. In Chapter 7, “Is Planning Tractable?”, the book discusses the type of hardware required to run software with these algorithms that could simulate society as-is; here, they also show that objections that the market is infinitely complex and cannot be simulated do not hold up to reality. Again, I largely agree with the authors’ perspective here, so I will not dwell on it.
However, in the book’s final chapter “Planning and Value” the gaps in the authors’ perspective shine most thoroughly. Thus far, the gaps had been glossed over, covered mainly by silences that did not make them explicit. The authors recognize that a way of valuing products, including raw materials such as steel and aluminum, is needed for planning (p. 156). They recover their idea of using labor values here, but the authors warn the readers that “an objection to measuring cost of production via labor values is that they don’t take into account non-reproducible natural resources” (p. 159). The authors then proceed to address this issue by “propos[ing] a measure of value that considers all factors of production, including both labor time and emission rights” (p.159). An example is worked through, where “emission rights” are added to the picture: the economy must not only hold to requirements on the amount of labor, it must also not exceed a certain threshold of emission. And voilà: the carbon emission problem has been correctly incorporated into our planning without compromising the labor-time solution to the calculation problem.
Here, we see one of the consequences of reducing the ecological crisis to the carbon crisis: after all, emitted CO2 is a measurable number on which quantifiable restrictions can be placed. And one could argue that scientists have already made some reasonable recommendations on how much CO2 should be emitted in the IPCC reports. What is needed is for society to follow them. But this is not the crux of the issue. The problem comes in the following four and final pages. There, the authors state several ways of making their planning methods more flexible. The first idea was already present in Towards a New Socialism: if a good cannot be sold, it is better for the government to sell it at a price lower than its value and to produce less next time, than to let it go to waste. And if there is too much demand for a certain good, it can be priced up temporarily, until the plan can adjust for society to produce more. They state that even if this mechanism “mimics to some degree the operation of a market economy”, it is still “perfectly consistent with planning of their production” (p. 165). Up to here, nothing too objectionable. But the authors then discuss how to extend this idea to incorporate the effect of emissions, proposing incorporating CO2 emissions into the picture and finding a way to mark up products which are ecologically unfriendly and so undesirable for society to produce. By doing this, they incorporate a very neo-liberal theory of ecological change. They end the book (p. 167) with this particularly striking note, which is worth reproducing in full:
High emission products will only be produced as long as people are willing to pay a high price for them. Since fewer people can be expected to be willing to pay a higher price, fewer high-emission products will be produced. This results in a shift of production towards low-emission products. Rather than simply reducing the economy’s overall output to abide by emission constraints, the composition of output is changed to emphasize green products.
So, to summarize the book’s solution: the ecological crisis can be solved by large-scale planning which incorporates CO2 (and other ecological limits) into the algorithm’s constraints, and we can move toward more ecological products by using pricing schemes which disincentivize their consumption. The first step is questionable because not all ecological limits are easy to state mathematically, but it is not the one which I object to on a fundamental level: we will have to regulate our social metabolism in some way which uses some numbers. It is the second step I object to fundamentally, which comes close to declaring a labor-carbon equivalent that completely buys into a carbon-pricing logic, and which I think defeats the entire purpose of the book.
In a way, this comes from the fact that the authors try to have it both ways: restrictions to use of resources do not invalidate their planning method and keep labor as an objective measure of value, but the authors still need to find a way to move society toward more ecological methods of production. They thus come to a simple solution: use pricing discounts. However, the obvious question arises: how does one decide how much to “cheapen” a more ecological product? Furthermore, it is not easy to solve competing stakes – how does one even begin to compare the appropriate labor-price difference between a production method that destroys a forest with one that pollutes a lake? The Austrians have a solution for this: the capitalist free market, and we know it does not work well. But the authors do not do much better; by modifying their labor pricing to include things beyond labor, we are moving much closer to needing a universal equivalent than by using the temporary adjustments in pricing of TaNS, which were applied on a single metric.
Indeed, this approach is a downgrade over the approach suggested in TaNS (p. 66), which is very clear that this cannot be simply an economic problem:
If a hydroelectric scheme is proposed that will flood a valley which is both a beauty spot and a unique habitat it is pointless to search for some economic formula that will decide if the project should go ahead. The problem is political, not economic. That is to say, a decision requires a deliberate judgement of priorities, and cannot be reduced to a comparison of simple numbers, whether expressed in labour time or money.
The authors were clear that regulating humanity’s metabolism with nature was a highly complex political problem in TaNS. It is baffling that the authors chose this alternative eco-pricing scheme in their new book, thereby giving ammunition to their critics who say they want an “algo-cratic” economy. One could speculate that once they have chosen labor times as an objective metric, there are not many more things one can propose which are quantitative and “objective” – but the authors should know better.
Enframing and Revealing
At the heart of the issue is that C&C&D’s approach to society, like that of large corporations, is one where economics is seen as a science of enframing,2 where we can basically quantify everything going on and, given this quantization, use mathematical tools that give us a North Star orienting us to the correct direction to travel. We can command and control society’s metabolism. In contrast to this, the liberal approach to economics is best seen as one of revealing: one where we are limited to observe, make notes, and slightly nudge the system when it does not behave as intended; where from the (self-interested) actions of agents, we get a societal ordering which we can gently pilot toward the best possible outcome.3 We manage society’s metabolism.
In science, both approaches are useful. Enframing is very common in physics, where through analysis, we can break up the components of things into their constituent parts and make precise calculations on these parts. Equations not only allow us to calculate the trajectories of planets and the charge of an electron, but of more complex objects such as cars and computers. Meanwhile, in sciences like ecology, the approach of revealing is taken: one can make models such as the predator-prey equations, which have strong assumptions and work in restricted circumstances, such as between hare and lynx in the Canadian tundra. But in more complex ecologies, such as those present in temperate regions, it is very hard to predict what happens when you add or remove a species from an ecosystem. One is left only with careful observation and gentle experimentation.
This distinction between approaches is nothing new. Even if all sciences fall somewhere between both extremes, since the scientific revolution and the advent of capitalism, the method of analysis and enframing has been the preferred approach to dealing with nature. This has been seen in either positive ways (the term scientific revolution itself has a high positive connotation) or in negative ways, by authors like Carolyn Merchant and her Death of Nature. This debate has seen many spins; it includes calls to incorporate “indigenous,” “synthetic,” or “holistic” sciences into the curriculum, or to solve the limitations of pure analysis which is unaware of its biases and tends to generate data which has racist and sexist biases in the less exact sciences.
I will not delve any further as it is beyond the scope of this review. What is of use here is the relationship to the calculation problem. In the method of “revealing”, this problem is solved from the point of view of a universal measure of value, arising through some sort of “myth of barter”:4 ever since the Stone Age we barter, and money is just the modern and rational way of doing it. We rely on individual people to adequately price things, and when they can’t, the state can step in to correct this “externality”: through carbon-pricing, fines on pollution, etc. The market, plus a few (very small!) nudges, will solve everything.
We know this doesn’t work, otherwise you wouldn’t be reading Cosmonaut. So, we need an alternative. What C&C attempt to do is to propose and impose a universal measure of value on society, based on a series of reasoned arguments, which I will quote at length (p. 158):
In Marxian political economy, the labour value of an item is the socially necessary labour time required to produce it. Given the technological, environmental and social conditions of a society, it is the labour time that is generally required to produce an item of this type. This includes the indirect labour that is necessary to produce raw materials, tools and so on. We can calculate the labour values of products if we have a detailed input–output table of the sort described in Chapter 6.
This step has been critiqued as a way of “naturalising” value and making it trans-historical.5 I do not think that framing helps; it can make the critique confusing. I tend to agree that some principle similar to “from each according to their ability” should be used in a transitional economy, and hours worked is a very fair one. A better way of phrasing the criticism is that by answering the Austrian critique with a different universal measure, the authors are imposing a choice which, even if it improves the distribution of wealth over capitalism, can constrain the system too rigidly. Furthermore, attempts to make it less rigid of an accounting mechanism will end up reproducing many aspects of market logic. In the TaNS system, we saw flexibilization number one: price discounts for under-consumed products (and the converse price increases for over-consumed products). In Economic Planning we see flexibilization two: price adjustments for desirable ecological products which mean we have to “account” (enframe) nature in some way, and through which we recover the logic of carbon-pricing. As we begin to address societal problems, more questions will arise. Would traffic fines be paid in labor-vouchers? Should we use differential labor-prices to regulate the cost of car parking in a city center? Yes, there are more complex methods to solve these issues, but the brilliance of capitalism is that everything can be solved through adequate pricing. And a cursory reading of the authors’ proposal without thinking deeply about these issues can lead us to thinking, “we will just solve this in a very similar way.”
Dense Organization as a Solution?
Let us be fair and take the other option. The authors left that pathway open for us in TaNS, as stated above. Let us use “more complex methods” to solve societal problems. After all, as Elinor Ostrom has well documented, small communities tend to solve complex problems rather effectively through self-organisation. In Governing the Commons, she includes ample examples of communities which were able to (self-)regulate their relationship to scarce resources provided from nature through the building of rather complex institutions – without being coordinated by “the market”. These include examples from farmers in Mediterranean Spain and the Philippines to communal tenure in Japan and Switzerland.
We can turn the dial completely towards the “science of revealing” end; we can say that the job of planners is just to provide adequate scaffolding to communities, and not to interfere too much with their internal operations. Society will figure out a way of using ecological and other limited resources adequately through deliberation. Planners are no longer the designers of an economic system, but the navigators of cybernetics. Planners do not control the economy, they manage it, like one tends a garden.
However, how does this “scale up”? Ostrom already mentions that these structures only work for a limited number of participants. Furthermore, how can we coordinate supply chains and utilize advantages of scale in this societal organization? It is not strange that many proponents of planning at this end of the spectrum end up proposing things like two currencies, or two exchange methods: one for local products and one for outside products.6 This would allow each social unit to self-organize in a way, building complex interaction pathways between their members to adequately use resources and prevent the destruction of ecosystems, while using a simple equivalent when trading with the outside. This could, in principle, work as a first bridge between countryside and city: the complex ecological pathways of agriculture need not be communicated to the outside, and only information like how many hours farmers worked need be sent to the “center”. The “center” could give loose planning suggestions rather than detailed instructions, and it could impose control if one community’s activities begin to harm those of others.7
This begins to look more and more like a nested system of the cybernetics type: each top level gives loose guidance and allows for the self-organization of the bottom level. In case of trouble, it steps in, imposing a regulatory framework similar to that proposed in Ostrom’s work. And the universal metric of labor time can be used for large-scale accounting.
This leads us to our final issue: can we have our cake and eat it too? Is universal labor time the adequate measure across all scales of organisation? This is what C&C&D seem to believe, and I am not so sure. Even in the most analytic of all sciences, physics, we have different equations at the level of a molecule, a person, the ocean… the random motion of millions of molecules leads to the emergence of something called “pressure”. It doesn’t make sense to ask what pressure is at a molecular scale. Is this also true for society? Can we impose a single universal metric across countryside, town, city, country and continent? This is where the authors answer with a resounding yes, especially in TaNS, which covers everything from the Commune to International Trade; I am not so convinced.
My main objection comes from the idea of institutional mismatch. By necessity, there will be communes or societal units in areas which can support a much lower population, or that are dedicated towards extraction and nothing else. Unless one takes the abolition of city and countryside to an extreme, no one would want to abolish the distinction between city and mine. These areas would have a much lower societal “energy” and would require very different institutional shapes from those present in large cities in temperate zones. Getting these institutions to relate to each other, or a low-population societal unit to relate to the “center” will be problematic. After all, the center would receive some sort of summary statistic such as “hours worked” or “tons of ore extracted”, without having detailed knowledge of the operations, and with much information remaining qualitative or localized. This system could very well end up falling victim to Hayek’s critique of planners not having enough information, or it could attempt to impose order and end up damaging the complex networks that have arisen. Furthermore, the state and the existing institutions would take very different shapes in some areas of the world than others. This would cause wide mismatches that could be catastrophic, as Stephen Bunker points out in Underdeveloping the Amazon. This is a formidable challenge which is often brushed under the rug: we tend to assume that the state will look more or less the same everywhere – but there is ample evidence to the contrary. And if the state is seen as something external, resistance will brew.
Aside from this, there is the can of labor discipline, and how workers have historically resisted accounting methods when they have been perceived not to act in their favor. From the black market in the USSR, through the way Bulgarian workers resisted informatized accounting in the factories, to the lack of worker’s enthusiasm with Cybersyn in Allende’s Chile, getting every worker to conform with the new accounting mechanisms will be another immense societal challenge due to the suspiciousness of people toward things they might not fully understand and the amount of educational resources it will require. While encouraging workers’ councils at the factory would go some of the way, it is disingenuous to think that the accounting methods would work well all the time, and enough flexibility would have to be provided.
However, I think there is still hope. The Incas governed a relatively heterogeneous land; they were able to mobilize the amounts of labor necessary to reshape nature by building canals while maintaining a society with relatively communal norms across a rather large extent of South America, the Tahuantisuyo. So, experience tells us it could be possible to do this at a world scale. But it is also a world-historic challenge.
I hope that this review is of help in rephrasing the problems of planning in a constructive manner, that goes beyond the usual thought-ending dichotomies of decentralization and centralization. We need both simple codes and complex networks. If they can co-habit stably is a very difficult question to answer.
- See, for example, Kyle Thompson’s Tools for Viable Organizing
- I am indebted to Andrew Pickering’s analysis of science in The Cybernetic Brain.
- One could develop this argument a bit further, stating that microeconomics is probably closer to the “enframing” side, and macroeconomics closer to the “revealing” side, and the chasm between these two approaches is the Lucas critique.
- The phrase “myth of barter” is taken from David Graeber’s book Debt. Graeber and many other authors show that primitive human communities did not engage in barter except in exceptional circumstances, and thus the emergence of money as something transhistorical is a myth.
- See, for example, “The People-Nature Relation and the Historical Significance of the Labour Theory of Value” by George Liodakis.
- See, for example, Colin Duncan’s proposal of LETS in The Centrality of Agriculture.
- A possible disadvantage is that this method could end up locally sequestering resources, knowledge, etc., in ways that might make transfer difficult.