What is the 'calculation problem'?

[Blanquist]

what does that mean? when was it first brought up? did lenin and co address it?


sorry i know very little about economics

[Kotze]

The calculation problem is about how a computer can never make a calculation as fast as people that shift tokens around in an interpretative mass dance :rolleyes:

Sometimes, it is also about how much of the data that would be very useful for central calculation isn't documented anywhere, but locked in heads. I wouldn't say that's completely untrue. There is the important question how we make it probable that people have incentives for providing truthful information. A few anti-communist extremists claim that, even if people wanted to provide that important local knowledge to the center, a lot of that couldn't even be communicated (just like when you know how somebody's noggin looks, that doesn't mean you can draw it so well that the drawing is as useful to a third person as identification help as you are in person), and that effect would happen to a degree that makes planning impossible, which I don't find plausible at all.

[daft punk]

Yeah, it's highly overrated by libertarian types, a theory that socialists could never calculate planning properly. In fact socialism would go off trial and error which is all capitalism does anyway. And multinationals do their own planning already. We would make it integrated, and not for profit, and democratic. Simples.

[Tim Cornelis]

Kotze attacks a strawman he created. Daft Punk does not seem to understand the extent of the economic calculation problem critique.

The economic calculation problem was first formulated by Ludwing von Mises in 1920. Mises summarised it as follows:


Without calculation, economic activity is impossible. Since under socialism economic calculation is impossible, under socialism there can be no economic activity in our sense of the word. In small and insignificant things rational action might still persist. But, for the most part, it would no longer be possible to speak of rational production … All economic change, therefore, would involve operations the value of which could neither he predicted beforehand nor ascertained after they had taken place. Everything would be a leap in the dark. Socialism is the renunciation of rational economy.

The crux of the problem according Von Mises is the lack of a common denominator (market prices) on which to base economic decisions. How do you determine whether to construct a railroad or an airplane to transport goods? How do you know what is most efficient? Mises argues that in a market economy we have market prices that determine what is most efficient, but socialism, as it lacks market prices, has no basis on which to judge whether a railroad or airport should be constructed.

Arguing that multinationals already plan is not a refutation of this. Multinationals plan in accordance with market prices, which would not exist in socialism.


And multinationals do their own planning already. We would make it integrated, and not for profit, and democratic.

But then you're exactly missing the point. Multinationals plan with profits, and if we make it not-for-profit planning we lack a rational basis, according Von Mises that is.

[daft punk]

Not necessarily. What I mean is multinationals allocate resources (plan) internally with no real reference to any market. Yes they obviously plan for overall profit, but some internal decisions are not referenced to any market.

As a tiny example, if you are a supervisor in a company, many of the decisions you make are just whatever you think 'should' be best, and another supervisor might do things differently.

Ok you could argue that a company that only employs crap supervisors could go out of business or whatever, I'm just saying in the real world ton of stuff is pretty random. Eg do we buy that new piece of kit or plod on and do the odd bit of overtime when required.

But I must admit it's something that interests libertarians much more than me, I trust our socialist computer programmers and so on.

[Tim Cornelis]

Not necessarily. What I mean is multinationals allocate resources (plan) internally with no real reference to any market. Yes they obviously plan for overall profit, but some internal decisions are not referenced to any market.

As a tiny example, if you are a supervisor in a company, many of the decisions you make are just whatever you think 'should' be best, and another supervisor might do things differently.

Ok you could argue that a company that only employs crap supervisors could go out of business or whatever, I'm just saying in the real world ton of stuff is pretty random. Eg do we buy that new piece of kit or plod on and do the odd bit of overtime when required.

Okey, but then you're begging the question is the internal organisation of corporations efficient? You answer affirmative, but is it really?

Individualist anarchist Kevin Carson wrote a critique of the internal organisation of corporations by using Mises logic to argue they are inefficient operations. So corporations themselves are also inefficient.


But I must admit it's something that interests libertarians much more than me, I trust our socialist computer programmers and so on.

I don't think many right-libertarians argue a computer cannot complete the necessary equations to make a proper plan. That is, the processing of information is not the problem, but acquiring the relevant information is the problem.

Also, I don't see how computers can be used in all instances. Computers only calculate, but cannot reconcile between two different materials.

For example, should we use 10,000 tons of steel, 5,000 bricks, or 6,000 tons of iron and 700 plates?
In a market economy we have a common denominator. The former alternative costs, say, 10,000$ while the latter costs 15,000$. But in socialism, how do we decide?
----------------
How do we solve this problem? We can't. Does this mean communism is impossible? Absolutely not.

First, we need to recognise that markets are not efficient either. While communism/socialism lacks a common denominator as the basis of rational decision-making, capitalism only recognises one common denominator in decision-making: profits. As such, in capitalism it is considered "rational" to let millions starve if it is not profitable. Moreover, market prices are always at disequilibrium, and are therefore always misleading and flawed. Capitalism also lacks a rational basis.

Communism can make a cost-benefit analyses using calculation in natura, labour hours anticipated to be expanded and how much will be saved, input-output tables (material balance accounting), taking into account externalities, etc. Naturally, we are also able to determine supply and demand by measuring fluctuation in demand as well as having consumers communicate demand to the relevant production committee or producers directly (or both). (while Mises denies socialism can ascertain demand).

This does not solve the issue, but it does evidence that

1. While we do not have a rational basis, we do have several indications we can use that help us approximate "rational decision-making"
2. Capitalism (or rather market economies) too lack a rational basis* so it's pretty much the pot calling the kettle black

*indeed, both internally (within corporations) as externally (prices, markets) capitalism is inefficient. Internally because corporations are centralised bureacratic and top-down and therefore inefficient.

[Nuvem]

All that really needs to be understood here is that:

1. Von Mises' Economic Calculation Problem stands as one of the only legitimate attempts by any bourgeois economist to recognize atheoretical fault with a socialist economy.

2. It was brutally proven wrong just over a decade after it was presented by socialist economic theory being put into practice- quite successfully. He never addresses the results of any of the Five Year Plans and moves on to the third phase-

3. Von Mises then went on to become a pure rhetoritician, abandoning all theory, intellectual integrity, or decency, becoming an angry, raving, crazed old man until the day he died. This is when he gifted us with such gems as "Planned Chaos" and "Planning for Freedom"- after his one potentially legitimate theory advanced against Socialism collapsed in the face of practice, Mr. Von Mises became, for lack of better description, a lazy asshole.

[daft punk]

omg, i thanked a stalinist!

but Russia did prove that a planned economy could work well, however in the long term the disadvantages of a bureaucratic dictatorship outweighed the advantages of the planned economy.

[robbo203]

This might help

http://www.cvoice.org/cv3cox.htm

[Kotze]

What I mean is multinationals allocate resources (plan) internally with no real reference to any market. Yes they obviously plan for overall profit, but some internal decisions are not referenced to any market.
Okey, but then you're begging the question is the internal organisation of corporations efficient? You answer affirmative, but is it really?

Individualist anarchist Kevin Carson wrote a critique of the internal organisation of corporations by using Mises logic to argue they are inefficient operations. So corporations themselves are also inefficient.That sounds like a pretty bizarre argument, link? Big corporations that use administrative decisions internally instead of AUCTION ALL THE THINGS come into existence inside the free market and outperform others. Highly flexible prices can impede foresight and frequent negotiations are costly, so the market mechanism can make free markets disappear.
I don't think many right-libertarians argue a computer cannot complete the necessary equations to make a proper plan.Im' afraid many of them are actually that dumb.
That is, the processing of information is not the problem, but acquiring the relevant information is the problem.Ahem, I already mentioned that objection to planning in my "strawman" post.

[robbo203]

omg, i thanked a stalinist!

but Russia did prove that a planned economy could work well, however in the long term the disadvantages of a bureaucratic dictatorship outweighed the advantages of the planned economy.

It annoys me no end when Leftists warble on about the so called "planned economy". All too often this boils down to a kind of meaningless mantra which substitutes slogans for rational analysis

All economies entail "planning". Even the most extreme free market version of capitalism is full of "plans". Enterprises formulate plans and these plans interact and impact upon each via the market competition.

The proposal to eliminate all these separate plans - or to "plan" the interactions between them - by assimilating them into one single mega plan is what is classically called "central planning". The idea is that inputs and outputs should be matched up within a giant Leontif-type matrix in a premeditated fashion with targets being handed down to production enterprises. Since modern production is highly interdependent this requires each enterprise to rigidly fulfil its target as other enterprises would be affected if they did not. Additionally this means there is the potential for enormous waste built into the system.

Not only is such a set up wholly incompatible with socialism/communism - since it implies hierarchical authoritarian enforcement - it is also wholly unworkable. It is simply not possible to centralise all the necessary data to enable the necessary calculations and material balances to proceed, notwithstanding advances in computer technology. More to the point, even if a fully worked out plan was created, it would be doomed from the very first minute it was implemented. Change is the only constant in a changing world, as they say, and even something like a drought or crop blight which wiped out part of the American Mid-west's cereal output would throw all the carefully worked out calculations into chaos

Mises and particularly Hayek made much of this in order to "prove" the unworkability of socialism. But of course, their arguments prove nothing of the sort. It is precisely because they identitified socialism with central planning that they were complete unable to see what is in fact the decisive counter argument to the so called economic calculation argument. This entails recognising that socialism would NOT be a centrally planned economy in the above sense but would be spontaneously ordered via a self regulating system of stock control. How this might work is discussed here

http://www.cvoice.org/cv3cox.htm

One last thing, Soviet Russia was NOT a planned economy in this sense at all. This is yet another myth that needs to be exploded once and for all.

The so called "plans" worked out by GOSPLAN and handed down to state enterrpises were a complete farce. Almost always the target that was planned at the outset e.g. X tonnes of steel, would be modified in the course of the planning implemetation period to make it appear as if the target had been reached. So the plan did not shape economic realities so much as was shaped by these realities. Sometimes planning targets were not even made available to enterpises during the period in which the plan was supposed to be implemented.

Also, contrary to impressions , the Soviet economy was a lot more de-centralised than is commonly supposed. A huge array of decisions were left to state enterprises - of necessity - such as technical ratios, employment and even wage levels. It would be more accurate to depict state agencies such as GOSSNAP as fulfiling the role of horizontal intermediaries between enterprises who of course were legally required to make a profit which would revert to the central state and part of which would be returned to enterrprises via GOSBANK as investment capital . Means of prpduction were commodites which were bought and sold between state enterprises and subject to legally binding contracts

The differneces beteeen the Soviet capitalist economy and western capitalism was thus largely superficial. All of the primary features of capitalism - above all generalised wage labour - existed in Soviet capitalism every bit as much as in the West

[eyedrop]

Economists can't calculate their way out the front door, so I wouldn't trust them with anything regarding calculating. There is a reason why investment banks hire physics/maths people, they have the mathematical tools to actually model it.

When reading popsci economicsbooks it's silly how they attempt to shoehorn obvious multivariable problems into simplified shit, by just assuming that one variable stays constant.

I don't understand why so much attention is given to hte opinion a theoretical dude instead of people that actually work with things similar (or actually the calculation problem) to the calculation problem.

[A Marxist Historian]

Yeah, it's highly overrated by libertarian types, a theory that socialists could never calculate planning properly. In fact socialism would go off trial and error which is all capitalism does anyway. And multinationals do their own planning already. We would make it integrated, and not for profit, and democratic. Simples.

It's just plain wrong. In fact, Leontieff input-output matrices are just about ideal for making the calculations for a planned economy, as Soviet economists tended to recognize.

It's actually a much cruder version of the more sophisticated and to the point argument of Hayek I think it was, that if you have centralized economic command from the top, it's impossible for any group of planners, no matter what geniuses they might be, to account for all the variables in something as immensely complex as an economy.

This is absolutely true, and one of the problems of the Soviet economy for sure. But the answer is simple and obvious, that for a planned economy to work properly it can't all be run from the top, and that you need democratic input into planning from the entire society of producers, as well as input from the consumers, the market.

As Trotsky put it, a workers state needs to combine "planning, democracy and the market."

-M.H.-

[robbo203]

It's just plain wrong. In fact, Leontieff input-output matrices are just about ideal for making the calculations for a planned economy, as Soviet economists tended to recognize.

It's actually a much cruder version of the more sophisticated and to the point argument of Hayek I think it was, that if you have centralized economic command from the top, it's impossible for any group of planners, no matter what geniuses they might be, to account for all the variables in something as immensely complex as an economy.

This is absolutely true, and one of the problems of the Soviet economy for sure. But the answer is simple and obvious, that for a planned economy to work properly it can't all be run from the top, and that you need democratic input into planning from the entire society of producers, as well as input from the consumers, the market.

As Trotsky put it, a workers state needs to combine "planning, democracy and the market.

-M.H.-


This would presumably be a capitalist economy you have in mind and not a socialist economy, since you seemingly envisage the retention of the market which is totally incompatible with socialism. Trying to run a market economy democratically, I would suggest, is a forlorn and futile hope

In any event, it is doesnt get round the fundamental problem of central planning and the notion of single mega plan for the whole of society - its inability to respond flexibly to change. Democratising decisionmaking will not overcome this problem and ironically may actually make matters worse in the sense in that it would add yet another layer of complexity

The answer lies not in tinkering around with the discredited idea of a so called "planned economy" but to develop a model of a non-market socialist economy that is capable of behaving in a self regulating, self ordering fashion. That is to say, has some kind of feedback mechanism. This is only possible in a society that is relatively decentralised and has instead of one plan, many plans, and instead of one decisionmaking centre, many centres,

[Kotze]

It annoys me no end when Leftists warble on about the so called "planned economy". All too often this boils down to a kind of meaningless mantra which substitutes slogans for rational analysisHow fitting that remark is, given the kind of "discussion (http://www.revleft.com/vb/super-computers-and-t154397/index.html?t=154397)" we've had before :rolleyes:

To sum up robbo's views:

1. Central planning can't work because of computation complexity, since some dead aristocrats outside the field of computer science said so more than half a century ago.

2. Central planning also can't work because when everything has to fit in an airtight way, a small mistake makes the whole structure crumble. And the reason central planning can't use safety margins anywhere is HEY LOOK OVER THERE A THREE-HEADED MONKEY

3. When I make vague assertions about Decentralized Plans That Interact In An Interactive Process™ that is some serious science, when you describe any adjustment mechanisms that means you are a pro-capitalism, because capitalism also has mechanisms that uh react to... things.

4. Rationing by labour vouchers is unnecessary (and also capitalist btw) because scarcity will disappear by star trek magic.

[robbo203]

How fitting that remark is, given the kind of "discussion (http://www.revleft.com/vb/super-computers-and-t154397/index.html?t=154397)" we've had before :rolleyes:

To sum up robbo's views:

1. Central planning can't work because of computation complexity, since some dead aristocrats outside the field of computer science said so more than half a century ago.

2. Central planning also can't work because when everything has to fit in an airtight way, a small mistake makes the whole structure crumble. And the reason central planning can't use safety margins anywhere is HEY LOOK OVER THERE A THREE-HEADED MONKEY

3. When I make vague assertions about Decentralized Plans That Interact In An Interactive Process™ that is some serious science, when you describe any adjustment mechanisms that means you are a pro-capitalism, because capitalism also has mechanisms that uh react to... things.

4. Rationing by labour vouchers is unnecessary (and also capitalist btw) because scarcity will disappear by star trek magic.

Sorry, but this garbled nonsense


I gave you what is the classic definition of central planning - meaning society wide planning, meaning a single mega plan which replaces (or assimilates) the multitude of plans that spontaneously interact -whether this be in the capitalist market today or in any realistic version of a socialist society in the future.

I think I made it absolutely clear that the problem of central planning has little to do with computational complexity and the capacity for modern computer technology to compute. It has more to do with data gathering but, even more so, with the problem of constant change

Literally speaking , yes, central planning in this sense cannot work for the very reason you suggest - " everything has to fit in an airtight way, a small mistake makes the whole structure crumble."

As far as I can work out what you are trying to say behind the distracting smart alec patter, is that the central planners should incorporate safety margins - or what might be called "buffer stocks" - into their calculations. Well, yes, the idea of a buffer stocks is a good one and is one that can be very usefully incorporated into a model of a socialist economy where the overall pattern of material flows is spontaneously generated rather than arranged in a some premeditated apriori sense.

To be quite frank Im not a 100% sure what you are trying to say, though. If you are advocating a priori society wide planning - even with due allowance being made for safety margins - you are still talking about a system in which precise targets are formulated and handed down in top down fashion for literally hundreds of thousands, if not millions, of products - from ballpoint pens to nuclear fuel rods. This does not get round the basic problems that have been identified with such a system and make it completely unfeasible

Nor do I understand your reference to "Decentralized Plans That Interact In An Interactive Process™ ". On the face of it, thats what Im advocating and Im saying quite plainly that the existence of some kind of feedback mechanism does not make such a thing "capitalist" at all. In capitalism, the overall pattern of production is not planned but is the emergent result of numerous market interactions. Simely the overall pattern of production in socialism will not - and cannot - be planned but in this case will be the emergent result of numerous asnd perforce separate plans driven solely by the criterion of producing for use

I dont say labour vouchers are capitalist. This is just a silly jibe. Im not in favour of rationing by labour vouchers ion a socialist society because frankly I consider the LV model of rationing to be administratively unwieldy and vulnerable to other problems too. There are other forms of rationing which I believe are much more straightforward and easier to implement. I think in a socialist society there is likely to be a dual system in place - one for free access goods , another for ratioined goods - at least for a while

On scarcity there are different senses in which one can talk about scarcity. In one sense we can certainly talk about scarcity as ineradicable in that it has to do with the nature of opportunity costs. Everything we do has opportunity costs. If we do one thing we forego doing something else. If we use X tonnes of steel for this purpose we forego using it for some other purpose. This kind of scarcity Ive never suggested can somehow disappear - magically or otherwise

Then there is a different kind of scarcity which is an empirical relation between the specific supply of ,and demand for, particular goods where both supply and demand functions are variable and adaptable . This in fact, is the kind of scarcitry that socialist allude to when they talk about socialism being essentially a post scarcity society

[A Marxist Historian]

This would presumably be a capitalist economy you have in mind and not a socialist economy, since you seemingly envisage the retention of the market which is totally incompatible with socialism. Trying to run a market economy democratically, I would suggest, is a forlorn and futile hope

In any event, it is doesnt get round the fundamental problem of central planning and the notion of single mega plan for the whole of society - its inability to respond flexibly to change. Democratising decisionmaking will not overcome this problem and ironically may actually make matters worse in the sense in that it would add yet another layer of complexity

The answer lies not in tinkering around with the discredited idea of a so called "planned economy" but to develop a model of a non-market socialist economy that is capable of behaving in a self regulating, self ordering fashion. That is to say, has some kind of feedback mechanism. This is only possible in a society that is relatively decentralised and has instead of one plan, many plans, and instead of one decisionmaking centre, many centres,

Of course a genuinely socialist economy won't have a market or market mechanisms. But, as you might have noticed by now from Revleft discussions, the pillar of "Trotskyism," insofar as that is a meaningful separate concept, is that socialism in one country is impossible.

In a transitional economy in between capitalism and socialism such as the USSR, or any other isolated workers state in a capitalist world, the market cannot be banished from the economy. Trotsky opposed Stalin's talk about abolishing the market and at one point even abolishing money during the so-called "Third Period," with its compulsory collectivization of agriculture--and its famine in the Ukraine.

Trotsky's follower Preobrazhensky, the economist of the Left Opposition, wrote some excellent work which has been translated into English. According to Preobrazhensky, the economy of the transitional period is characterized by the struggle between the market principle and the planning principle.

So what Trotsky advocated was input into the democratically discussed and debated central plan from below at every levels, from the factory floor on up. And that, given that you still have products produced sold in stores rather than distributed according to rationing cards as under "War Communism" or the height of the Stalinist industrialization drive, supply and demand in the market would help determine how many toothbrushes to produce next year etc.

In fact, with the entire working class involved at every level, and democratic decision making, you would have far more flexibility and ability to change the plans as change was necessary than in the Soviet system.

As for a situation which you project with a decentralized economy with not one but many plans, that would be extremely inefficient, with each area trying to produce everything everyone needed, instead of, as in a rational world, stuff produced in area X being sent to area Y and vice versa.

In fact, doing this was exactly one of the major problems of COMECON, one of the major reasons why Eastern Europe fell behind Western Europe, where the economies were and are partially integrated, first under the EEC and now in the EU.

-M.H.-