EconLog small logo  

Second Best and Socialist Calculation

MRV watch... A Cantabrian Commerce Centenar...

The socialist calculation debate of the 1920s and 1930s centred on whether a benevolent social planner could allocate resources in a planned economy to replicate the efficiencies of the market while remedying the ills of monopoly losses and profit-taking. Ludwig von Mises argued that without a market in consumer goods, one could never impute back prices to capital goods and consequently capital goods could never be allocated rationally in a planned economy. F.A. Hayek argued that the planning agency could never acquire the information necessary to form the optimal plan -- most knowledge is held in decentralised form that can never be aggregated to any one body; only the price system can work to that end.

The calculation debate was framed entirely in the context of benevolent planners. And, when planners are benevolent, it's very clear that planner inability to engage in economic calculation reduces welfare in a planned economy. But what if the planner isn't benevolent? Would outcomes in the Soviet Union, China or Cambodia have been better if Stalin, Mao, or Pol Pot been able to solve the Mises and Hayek problems?

Andrew Farrant and I argue (working paper version) to the contrary: we rather argue that the calculation problem raised by Mises and Hayek can never worsen and will sometimes improve outcomes in the socialist state. Why? If planners are benevolent, they'll retreat from planning when they see the problems raised by the calculation argument, as did Western Europe and Britain in the late 40s and 1950s. If planners are not benevolent and rather are seeking to operate as an Olson bandit, the calculation problem serves to limit their rapacity. In the limit, perfect ability to engage in socialist economic calculation transforms the planning bureau into a perfectly price discriminating monopolist in every market, extracting all surplus for the planner. While the calculation problem creates deadweight losses, it also ensures that at least some surplus is left for the hapless citizenry. Median citizen welfare improves even though total surplus is less than it would be absent the calculation problem.

Consequently, it is wrong to view economic calculation as the problem with planned economies. Yes, inability to engage in economic calculation ensures a fair bit of economic irrationality in socialist states. But, it also constrains non-benevolent dictators. As Bryan Caplan argues, incentives, not accounting, lay at the heart of socialist failure.

Comments and Sharing

CATEGORIES: Austrian Economics

COMMENTS (15 to date)
liberty writes:

1. There is no clean divide between "benevolent" and "non-benevolent" planners. Marxists believed that they would bring the best society possible into existance through planning; via socialism they would reach communism. Communism would make any sacifice worthwhile. If planning wasn't working, they may temporaily retreat (eg the NEP) but they would never stop trying. You can then label all Marxists as "non-benevolent" but they certainly considered themselves to be benevolent. You will have to define after the fact any believers in planning as non-benevolent because anyone who stops planning and returns to capitalism surely is not a true believer.

2. In the USSR after Stalin, planning was not done by a singular entity; it was well distributed and beurocratic. Were those planners benevolent or non-benevolent? Surely they were just trying to forward the country. All of them could see the inefficiencies and problems with calculation - they wrote about it especially starting in the 1960s in books to fill a library - but they did not consider radical reforms, most of them; certainly not the return of a market system; but they instead attempted to use small reforms within planning to improve efficiency.

3. As for non-benevolent planners being limited by the problem of calculation - I don't think so. Rather than limit Stalin, it gave him an excuse to use more totalitarian measures and the same was true with Mao. The best way to ensure success of a planned economy riddled with calculation problems is to focus the economy on the most important few sectors of the economy (using "shock tactics") or building a military-style superstructure and enforcing planning directly from above. The calculation problems were used an excuse by the non-benevolent to kill the saboteurs - difficult to prove that it was calculation and not sabotage that caused famines; they were used to make planning more hierarchical and dictatorial; they tended to lead to more brutal and totalitarian states under the non-benvolent rather than any state which would lead to a higher citizen welfare.

4. Calculation problems intertwine with incentive problems, for example because of calculation probems some decentralization was introduced in the Soviet system and obyedineniya were given more technical power to determine how to fulfill plans. Calculation problems meant that planners could not tell the firms exactly what to do, they could not know details (Hayek) and they could not know what costs or demand were (Mises), so they had to simply provide a plan via a single target and incentives via various cost-based guidelines. However, no matter how decentralized the system got, the managers at the firm level would still have to fulfill a plan not maximize profit (otherwise the return to market system would be swift). So, the firm would be given a plan target, such as 200 tons of boots; and then the incentive structure thereby created (even offset by the various additional incentives provided by the Gosplan) meant that they would choose the heaviest and cheapest leather and make only one size of boots. Transport firms would drive in circles with heavy loads in order to clock ton-km, etc (Nove). If costs, demand and details of the output could have been known, the firms could have been given detailed plans saying "buy 100 tons of medium thickness leather from Firm XYZ and use metal tips from Firm QRS, make sizes 32-50, ladies, in red ..." but since they could not know, they could not provide detailed plans and this skewed incentives much further than lack of profit alone. Skewed incentives meant possibly less work for the workers, but also worse products for everybody and lower output resulting in lower standard of living for everybody.

Scott Scheule writes:

Damn, liberty, well said. Who the hell are you?

liberty writes:


Thank you Scott. I am just a student of economics with particular interest in this area. Check out my website.

[URL fixed to read --Econlib Ed.]

Eric Crampton writes:

Liberty -- you really figure that if Stalin had all the information to form the benevolent plan he wouldn't have used it to extract maximally? Yes, consumer goods would have meshed better with consumer demand absent the calculation problem, but each person would be charged his or her reservation price for it....

liberty writes:


If Stalin had been able to plan without calculation problems, he would have been able to support the military and industrial buildup that he desired *and* provide subsistance for the masses. He would have extracted the reservation price from all consumers, but they would have still been better off than in the situation that prevailed where agricultural famine and misallocation of resources meant that most people were not able to purchase goods at *any* price.

(-_-) writes:


How would Stalin aquire this information? (yes this is Hayek again)

The most obvious answer to the benevolent information you speak of is capitalism. Unfortunately, in capitalism, Stalin has no power...
I am not sure I am contributing anything to this blog, Liberty summed it up very well.

Eric Crampton writes:

(-_-): I'm not saying he could; I'm saying it's a good thing he couldn't. Inability to engage in socialist calculation improved outcomes in socialist countries relative to the alternative where the non-benevolent planner extracts all possible surplus.

Liberty: You're assuming that the famines were in error rather than deliberate. If they were in error, then you may be right: ability to engage in calculation would prevent that. But if Stalin did it on purpose to starve out the kulaks...different matter entirely. Recall the definition of a reservation price as well -- that's the price at which you're indifferent between having and not having the good.

liberty writes:

Since it only makes sense to actually read the full working paper and address critical points made within; I have done so. I apologize for repetition and length. I should probably write it up in a well researched paper and post it on my site instead, but I will post this here now and perhaps follow up later with a better written critique.

The problem I see in your use of "benevolence" remains that you can easily determine afterward who was benevolent and it will fit your model. But Lenin by all accounts, agreed by nearly all historians and by the mass of his writings, truly believed that communism would be the true savior to the world. As did all Bolsheviks, Mensheviks and many others at that time. Nor did they give up as soon as "war communism" proved a disaster and famine spread across the countryside. They half-retreated to the NEP and all true believers at that time - even those who did not seek any personal gain whatsoever, political or otherwise - were determined to again find a way to build socialism. Far from recognizing the strength of the Mises' critique (granted, not yet written) -- or their own personal experience with the inability to plan or for workers to plan themselves at a decentralized level -- they were adamant that their partial retreat was temporary and that they would soon figure out how best to build socialism.

Many of those well known thinkers engaged in the industrialization debate during the twenties were vying for power, but the debate was not simply a veiled struggle for dictatorial control; it was a true ideological and economic debate and it went well beyond the limited most-read circle of Trotsky, Preobrazhensky, Bukharin, Stalin and so forth. The Institute of Red Professors was involved with students and professors allying with one or another of the major thinkers and writing papers and pamphlets of support. Communism was ideologically driven toward a return of central planning that would work this time. All of those individuals involved were not vying for power and driven by self interest. The planning produced a need for dictatorial control - and the problem of economic calculation was not innocent in this regard. The difficulty of calculation made it that much more necessary to centralize the system; build large collective farms; repress private trading, etc.

Next you do admit that the benevolent may remain advocates of socialism - so long as it isn't because they believe in its efficiency.

Your assumption that the calculation problem improves consumer welfare rests on the assumption that the dead weight loss to the consumer under imperfect calculation is *less* than the loss of his consumer surplus to the monopoly state producer which leaves him at his reservation price but able to purchase the goods.

In perfect information planning (impossible but theoretically) the consumer could purchase goods so long as he was willing to pay his reservation price. Given that his reservation price is being manipulated by a monopoly producer of *all* goods, this may leave the consumer with only "necessities" and little else. The "non-benevolent" planner may then use the excess to buy palaces, build armies and dine on caviar in foreign states.

However, this situation may still be better than the situation where lack of calculation is added to the mix. Now, rather than all goods being available for the reservation price; goods are not available and are made poorly and one must wait in line for them. How can we know that it is worse? Because in the latter scenario: the non-benevolent planner still pays the least he can and charges the most he can; but now he does not know how much to make of any given product, incentives are further skewed so that fewer products are made than even planned, are made poorly, less is produced and what is produced is often not distributed. Calculation, while not the only factor, is a major factor in each of these; the non-benevolent planner now can extract all surplus of a smaller pie: leaving the people with even less.

Next you make the assumption that an inability to know which effective (if indirect) taxes will cause inefficiencies, and thereby reduce the ability of the planner to extract all the surplus. This does not match the historical reality though, as prices were often set without regard to creation of inefficiencies. Sometimes, its true, planners would raise wages in order to induce labor into a new arena - but even the fully informed planner would have to do this. There is little reason to beleive that the less informed planner would have to raise them higher as the wages were set on trial and error basis. Your example of Stalin and marginal tax rates is enlightening - but it does not prove that the productivity loss of lower calculative efficacy is made up for by any difference in taxation that might have occurred. Stalin may have been very close to correct, and the productivity difference due to reduced efficiency may have been huge.

Again you assume retreat on the part of the planner if efficiency is so low as to cause famine: what of Lenin, Stalin, Mao and Kim Jong Il then? Lenin only partly retreated and planned to return; Stalin did not retreat at all; Mao did not retreat until others forced him to at peril of his life and Kim Jong Il still has not retreated after decades of famine.

A rational, wealth maximizing despot might retreat -- but what if the planner is either not wealth maximizing but ideological (Lenin, Mao) or partly-maximizing, partly-ideological (Stalin) or a crazed lunatic (Kim Jong Il)?

The problem with assuming rational wealth-maximizing behavior on the part of a single planner (or even a small group of planner working as one unit) is that economics doesn't work that way: not every single individual is completely rational and maximizing. We make that assumption because as a whole people behave that way. When the shop raises the price on tofu, some individual vegetarian might actually buy more because she wants to support the company. Its ideological and irrational and doesn't fit the model; but she is only one person making only one purchase so the model still works. But when Lenin makes choices like that, it messes up the model if he is the sole planner.

Then you admit this by saying that the non-benevolent planner *may* choose to allow starvation -- something much more likely to occur under a low calculative effeiciency - once again showing that economic calculation *is* important, with a non-benevolent planner. You suggest that with better efficiency it would be easier to induce famine - perhaps. But Lenin did not mean to induce famine and yet let it last for some time because it was poltically feasible and in some ways helped to serve his ends while ending it would mean retreating, an obvious political impossibility; same with Stalin; same with Mao. Any of them may have avoided famine if they had better economic calculative ability. Why would Stalin induce famine on purpose if he could collectivize without it? His purpose was collecivization and the buildup of the industrial state and the power over it; if he could do it without causing famine, it would have made him look better and he would have had less to worry about politically and economically - he would have been able to create more wealth for his military and more soldiers for his army. etc.

You assume that only stupidity would allow a planner to not retreat from planning: Stalin was not stupid. He allowed famine for a certain amount of time, but improved efficiency and did not retreat from planning.

Finally, planned economies are shortage economies; they are not demand driven, they have resource constraints not demand constraints. Shortages drive the economy, rather than demand (Kornai). This is not due to calculative inefficiency - in a perfect information model, it would imply perfectly efficient planning with full employment and full use of all resources. However, with lack of information and ability to calculate, prices can be set at any level that a planner desires and shortages will still occur because they will occur at various points in the production process for each of the goods in the economy and generalized (and self-breeding) shortage will result in limitation of use of wages on any basket of goods, as goods are in short supply. Generalized shortage results in spending 10 rubles plus half a day's wait (when the price is set at 10 rubles) instead of 50 rubles which the fully-informed planner might have charged. This implies that consumers only get their reservation price - at best - regardless of benevolence in any economy with scarcity and imperfect information. So, this would suggest that it is impossible to improve the situation with a reduction in calculative efficiency.

Eric Crampton writes:

- Benevolence means that the planner cares only about making the citizens better off.
- The retreat to NEP happened when the Communists' power was threatened
- Consumer surplus under the perfectly extracting planner is zero. That's what perfect extraction means. The median citizen must be made better off by reducing planner calculative efficacy where the planner is non-benevolent.
- If the planner isn't benevolent, there's no reason to expect retreat in face of famine if famine suits the ends of the planner.

Matthew Cromer writes:


This is the best guest-blogging I've seen done anywhere. Kudos!

Bryan, congratulations on getting such a high-caliber contributor to keep up the standards here while you are gone.

liberty writes:


Benevolence means that the planner *wants* to make the citizens better off but a) it is impossible to know whether any individual planner wants that and certainly whether a group of planners wants that and b) they may want to make the citizens better off by bringing socialism. There are not only two options - they may neither want to maximize their own wealth nor do whatever it takes including a retreat to capitalism in order to maximize the citizens wealth: few historians would agree that either of those two descriptions would fit Lenin, for example.

Not sure what you mean by that second point: do you consider that retreat a sign of benevolence or non-benevolence?

Simply because consumer surplus is zero under perfect planning does not mean it is higher under inefficient planning: inefficient planning leads to reduced growth; the total productivity is lower and the zero-consumer-surplus is applied to a different level of output. People go from subsistance and the barest ability to spend all earnings on the least supply the monopoly power is willing to provide for that amount of money to LESS defined by LESS total output and shortages of even the barest essentials.

The last point we have already addressed - however the planner may not face a famine in the first place if there were no calculation problem, in which case he may not choose to create one - for political or other reasons, not because of benevolence.

But I have made these points already. Apparently we do not agree.

Max U writes:

Liberty writes:

>Simply because consumer surplus is zero under perfect planning does not mean it is higher under inefficient planning: inefficient planning leads to reduced growth; the total productivity is lower and the zero-consumer-surplus is applied to a different level of output. People go from subsistance and the barest ability to spend all earnings on the least supply the monopoly power is willing to provide for that amount of money to LESS defined by LESS total output and shortages >of even the barest essentials.

I am unsure what on earth Liberty means above in terms of very basic micro, but it is hard to get much lower than zero (Simply because consumer surplus is zero under perfect planning does not mean it is higher under inefficient planning)

Would you rather have 100 without Stalin (the first-best); 50 the second-best; or 100 - which is then taken by cuddly Uncle Stalin.

If the first-best is not feasible, then I choose the 2nd please.

liberty writes:

>I am unsure what on earth Liberty means above in terms of very basic micro, but it is hard to get much lower than zero (Simply because consumer surplus is zero under perfect planning does not mean it is higher under inefficient planning)


Zero consumer surplus does not mean zero output consumed - it means the very least consumption that a person would be willing to work for.

So here is how. The way the consumer surplus and producer surplus are divided have nothing necessarily to do with the original output created. If GDP is 20 million rubles and then Stalin takes 90% of it because as a monopolist he can, leaving the consumer with only the lowest wage and highest prices possible - zero consumer surplus - then the consumer base has still managed to accumulate 2 million rubles worth of goods. The people would not work for less and are paying tha maximum they are willing to pay for goods, but they still consumed 2 million rubles worth of goods - and Stalin got fat on the other 18 million.

In an economy with poor calculation, it may only create 6 million rubles output. In that economy, Stalin is unable to extract 90% of the output in producer surplus, he may be forced to allow the consumer some surplus nominally - however, the higher wages cannot purchase goods because there are no goods to buy and everyone must wait for hours or days on line to get what is available. So, economic calculation problems mean that Stalin can only extract 70% of the output for himself; yet consumers are left with only 1.8 million rubles to consume -- which is less than the 2 million under perfect calculation.

Eric Crampton writes:

Liberty: we're using terms to mean different things, methinks. In our model, the consumer has only epsilon utility: just the bare minimum necessary to keep him from commiting suicide. Every ounce of surplus has been extracted from him by the planner when the planner has the information necessary to enegage in economic calculation. All consumer and producer surpluses have been transformed into planner surplus. It doesn't matter how much money's worth of goods consumers have then consumed. They're still just barely preferring life over suicide.

Think of it this way: which is better for you as a consumer.

A: You earn $10. There is one good on the store's shelf. You're just barely willing to spend $10 on it. If it were $10.01, you wouldn't buy it.

B: You earn $5. There is one good on the store's shelf. It isn't a very nice product. You'd be willing to spend $6 to buy it (if you had $6), but the price is only $5. You buy the good, and have $1 in consumer surplus.

In your story, folks are better off in A because they've got more money's worth of stuff. In mine, they're better off in B because they still have some surplus.

liberty writes:

I'm not sure how your model shows that the consumer is left at committing suicide level; it seems more to me like you've shown that they are at the level where they are barely willing to work. All surplus is gone because the consumer's entire income must be spent to get the barest essentials required so that he will show up to work the next day.

When the calculation is imperfect the planner may want to extract all surplus and put the consumer at that level - barely willing to work - but the economy cannot generate and distribute enough goods for all consumers to be sustained at that level.

Comments for this entry have been closed
Return to top