September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« M. June Flanders on Horwitz vs. Wray | Main | Cobden Centre Picks Up Whatever Happened to Efficient Markets »


Feed You can follow this conversation by subscribing to the comment feed for this post.

More support for saying the Samuelson obits were far to respectful, for a man who has a lot to answer for.

A good bio might help us understand the pathologies of this man and his place in economic thought.

The comments on this at Alex's blog lay the blame on Samuelson's models, and point out that an even more left-wing econommics text was more skeptical of the Soviet economy. But for Samuelson not to have been at least suspicious to the point of redoing his "model" (or whatever it was) leads me to conclude that ideology played at least some role in his blindness. I agree with Greg's comment--the man had much to answer for. For someone who was supposedly the sharpest guy around, he didn't know a lot. (Sorry, Yogi Berra could say it better.)

"It has long been my view that the mismeasurement of Soviet economic growth is just the tip of an iceberg of methodological and analytical malpractice in economics that affects not only our understanding of the economic history of the Soviet Union, but also the plight of the less-developed world and also the reality of economic life even in the market economies of the West."

I think this spot on - and I think the flip side of this is that a close analysis of why the Soviet economy could not be measured in the same way that market economies *supposedly* can, and of the inefficiencies that were missed and that are missed in aggregative measuring techniques, and of the alternative ways to measure, would be very useful for the economic science.

Economists should also study the models that they had been using which told them that the Soviet Union was a raging success - they should ask themselves what in their model was wrong, what was lacking, why did the model fail to show inefficiencies? Then they need to ask themselves whether it might miss inefficiencies in mixed-market economies too.

Economists should ask themselves whether they are OK with the fact that their models told them the Soviet Union was a raging, roaring success, and in fact it was a crumbling, inefficient and chaotic mess. They should ponder whether it might indicate that (a) they should really start from scratch on the modeling front, and (b) perhaps they should go visit the economy they are modeling some time and take a moment to meditate on whether their model captures the reality at all.

It is pathetic to me that most economists have failed to do these things - the economists at this blog excepted.

Most of this was beaten to death here in an earlier round. I shall simply repeat that I think the biggest problem with the measures of the Soviet economy was the inability/failure to measure how poor the quality of their goods was. The numbers on the amounts of things they had were not all that far off, although there were certainly some cases of that, e.g. the cotton factory in Ubekistan that was supposed to be there that was never built. But those CIA and GOSPLAN numbers on the amounts of steel and cement and wheat and tanks and so one were not all that off. It was only when things were opened up and it was learned that steel made at the Lenin Steel Works in Magnitogorsk could only be sold as scrap metal that the nature of the problem became clear.

Barkley - yes we have beaten this a bit to death, but you still insist it was only a matter of quality, when in fact the aggregate data hide much more inefficiency than just that. You confirm your own mistaken assumption by looking at the aggregates - looking at the aggregates after the opening of the archives and seeing that they are similar to the estimates before is not enough. The aggregates hide the truth.

In addition to the problem of low quality, there were also enormous amounts of unsold inventories (which, as with everything else, was known by some if not by the CIA by 1965). This was only partly due to quality issues. It was partly due to NOT KNOWING DEMAND and NOT BEING ABLE TO COORDINATE - i.e. making too many of one input and too few of another.

There were also endless investments that were unfinished, dragging on for many years, often finally abandoned.

A few examples:
"During the first four years of the seven-year plan, unsaleable stocks of textiles, clothing and shoes more than doubled; they increased on the average four times as fast as sales."

"Shocking errors in planning appear side by side with a further aggravation of deficiencies that have existed for a long time. Thus the amount of investment funds “ frozen” in unfinished projects assumed increasingly dangerous proportions in the course of the last years of the Khrushchev period. Each year between 1958 and 1963, additional billions of rubles were “frozen” without any kind of “return” whatsoever; their total amount reached 25 billion rubles by 1961 and passed the 27 billion figure by 1963 (these two figures represent 75% of the respective total investment expenditures for the two years in question)."

"In 1963, for example, the USSR produced 206 million tons of crude oil, but the total annual capacity of Soviet oil refineries reached only 50% (!) of this level of production."

These just from one report:

Aggregate data such as GDP does not take into account whether the output is actually being used or whether it is sitting on a shelf! It counts half finished development and intermediate goods!

This is the kind of thing that economists need to start realizing about their statistics and models!


I do not think you are disagreeing with me. You agree with me that in fact the aggregate numbers on production ("they had" as I put it in my comment) were not all that far off. You accurately bring up that they had an inventory problem, which should be accounted for. However, I will contend that the inventory problem is at least an order of magnitude less significant, if not two orders of magnitude, as a source of error than the quality problem. It is a minor sideshow. If you want to bang the drum on coordination problems of central planning, well, duh, I made no claim that they did not have such problems, and for that matter, Samuelson from very early on emphasized the micro inefficiencies of their system, with long lines of varying lengths with resulting misallocations, blah blah blah. There is no debate about that by anybody here or even by Samuelson himself.

Let us go back to the German case. The CIA was estimating in 1989 that the real per capita income of East Germany had just about caught up to that of West Germany. After the wall fell, the reestimate was that their real per capita income was about a third to maybe 40% of that of the West. This reestimate was not due to piles of unsold inventories of crappy shoes and Trabants. It was due to just how crappy the shoes and the Trabants were. The inventory issue probably accounted for 1 or 2% of this reestimate, totally insignificant (oh, maybe 3%, if you really insist). Sorry.

BTW, if you go over to MR you will see that I say nice things about you, but here you are making yourself look silly. This is a sideshow.


Could you provide a cite for the assertion that the inventory issue only accounted for 1-3% of the re-estimate? I would add - the re-estimate may not have been perfect either (!) and may have missed many of the coordination problems itself for many of the same reasons that the CIA missed them.

You may say I look silly for pressing this issue, but I would be very surprised if these coordination problems were as slight as you say. Just the single issue of rubles frozen in investment projects yielding zero return and finally abandoned was a couple of percent of the total economy. Add to that the fact that 50% of the oil extracted could not be refined because there was no capacity, then add to that all the "hoarded" resources that were not in use, all the truckloads of goods sitting idle -- certainly Stalin and Khrushchev would be pleased to learn that these only represented a couple of percent of the waste in the economy, because according to their own estimates the waste was much larger than that, and the resources they used to try to resolve these problems themselves were in the many tens of billions!

The 1965 reforms addressed these issues primarily - although the issue of quality was also a significant one. Why do you think the leadership would undertake a massive reform of the bureaucratic structure, and spend many meetings and thousands of pages of articles on issues that only amounted to about 1-3 percent of waste in the economy?

I may also add that - although I do not trust his estimates for reasons I discuss in my book (essentially, he could not estimate much better than planners could estimate) - Noam Jasny attempted to estimate the waste due to calculation problems due to centrally set prices in the Soviet economy.

Some of his conclusions:

"Altogether the cost of all power produced in the USSR may well be one-third or more larger than it would have been if the choices at the time of construction were rational."

"Round timber is shipped in great amounts to consuming areas over particularly long distances; the local sawing facilities are utilized only little (20%). As the reason, the faulty structure of prices and of the charges on transportation of round and sawn timber is given.
The fact that such round timber shipped over very long distances still finds consumers (these possibly are getting the timber by allocation) shows incidentally the confused state in which the market of this important building material is permanently found.
The cost of timber used in construction work is not much under one-half of the total cost of building materials used in the USSR; so the losses due to mismanagement in the timber industry run to quite a pretty penny."

"In the USSR, however, the ratio between the prices paid to kolkhozy by the State for pigs to the price for feed grains is I4:1 or even higher. ... This 14:1 ratio is the result of an agglomeration of inefficiencies. There is, for example, an immense amount of waste of feeding stuffs"

"One can but hope that the immense extent of the losses caused by irrationality and inefficiency emerges clearly from our analysis. The reader should bear in mind that the source for covering the losses is also immense. It is the underpayment of labour; a hundred million persons are toiling for rewards which, in the most modest reckoning, are a quarter or a third less than what they should be."

A Note on Rationality and Efficiency in the Soviet Economy. I A Note on Rationality and Efficiency in the Soviet Economy. I & II
Naum Jasny
Soviet Studies, Vol. 12, No. 4 (Apr., 1961), pp. 353-375
& Soviet Studies, Vol. 13, No. 1 (Jul., 1961), pp. 35-68

"The main political and economic problems of the Soviet state revolve around the question of prices."
Institute of Economics of the Academy of Science, USSR


I do not wish to get into another round of rehashing all this here. Tired of it and way overswamped with piles of stuff on my desk after the AEA meetings. However, if you want to throw your book at me, please tell me offlist who the hell you are and what your book is. I think you know how to find me.

The point about the coordination problem is not that it was not a problem. It is that it is one that most of us knew/know about, although it may be argued that Samuelson and others somehow did not properly account for it in terms of the GDP estimates and so on. But, again, they were basically adding up production, so coordination should not matter for such a count unless one gets outright lying about numbers (which did occur). If oil was not getting refined due to a refinery shortage or problem, then one would not see refined oil showing up in output counts. This is really a sideshow for the measurement issue, even if it is profoundly important regarding fundamental evaluation of the system.

I do not have source on the inventory number, but this is just simple. To alter the GDP by more than half, or even more than 10%, one would have to have acres and hectares of storehouses for all the unsold shoes and Trabants and so on and on and on. Sure, those inventories were piling up in the USSR in the 80s, but the mismeasurement goes to much earlier. You want 10%, maybe that is true for some country or other, but certainly not East Germany. There are lots of references on how shocked people were when the market for some of the crummy stuff just collapsed to near zero overnight, and the listed value of this output also collapsed. This is well known to have happened, and the decline fits well with the larger decline in estimated GDPs. There is simply no way that the inventory issue was going to be larger or even anywhere near as large as the quality adjustment issue.

Hey, why are you so down on the quality point? This is clearly a deep problem for the system, and one that is in the long run much more telling than some inventory problem. Indeed, it is the fundamental reason for the inventory problem. People were not buying the shoes because they did not fit and the authorities did not bother to make them fit the sizes wanted by the public because of the lack of market feedback on production.


My signature links to my website - we've met and chatted before (my name is Guinevere Nell). I am not "down" on the quality question, I am simply making the case that the calculation/coordination problems were indeed severe. My focus has been the Soviet economy, not the East German economy, but clearly many experts consider the calculation and coordination problems - plural - to have been very serious indeed. As the Jasny quote states, it may have reduced GDP per capita by up to a third. As the Institute of Economics of the Academy of Science quote stated, it was the primary problem facing the Soviet economy in the 1960s. There are many many more quotes by Sovietologists, Soviet political leaders, Soviet economists and others to the same effect.

Yes, quality was a major issue - but quality was also affected by pricing problems, and coordination and calculation issues that came from them, as targets replaced profit as the guiding force, etc. My only point is that it was not quality __alone__ that was affected. Many other problems were caused by the calculation/coordination issues that stemmed from lack of market prices, and planning.


Never linked (and it is blocked on my computer now), but indeed we have met. Saw your dad at the AEA meetings in Atlanta.

Regarding the quality issue it is neither a pricing nor a coordination issue, although it could be, especially the former. It is more the incentives issue. If firm managers (or whoever is making the real decisions) does not get rewards from pleasing the market through higher sales, then there will be insufficient incentives for quality improvements. This failure to improve quality built up substantially over time.

Russia certainly had it, but the German case is in some ways more dramatic for the same reason that comparing the two Germanys was always a favorite thing to do for comparative economists, their similarities in so many ways besides their respective economic systems. So the sudden and stark contrast when the Wall fell and the two subsequently reunited showed this problem in an especially stark way.


my site is: in case you are interested.

"Regarding the quality issue it is neither a pricing nor a coordination issue, although it could be, especially the former. It is more the incentives issue."

Quality issues in the Soviet Union were indeed very much about not having to, and not being able to, serve the customer - but instead having to follow a set of targets to fill the plan. They were also very much caused by coordination issues, and directly pricing issues (which are themselves the cause of coordination issues). And finally, due to the distorting effect of targets - which is a planning issue, but not precisely an incentives, calculation (pricing) or coordination issue I suppose.

As you know, targets caused distortions in what, how and how well products were made. The dimensions of a product were chosen to fill targets, the types and attributes of the product made, and the way it was made were on this basis.

This meant two things occured among producer firms: (1) Heavy, thick glass and metal were made to fill output targets faster, etc. Hence many factories that *might* have produced half-way decent final goods were stuck with inputs that made certain the final good would be unsaleable in a market economy.

(2) But this also meant that many firms *did not produce* many of the intermediate goods and inputs that other firms needed (causing a coordination problem) - for example famously producing a smaller number of big house-nails, instead of a large variety of nails needed across the economy (inspiring the Krokodil cartoon of one humongous nail). Hence, there were also factories that simply could not acquire the inputs they needed, and had to make bizarre substitutes, reduce the available product choice they would offer, or leave off part of the design or construction, because they could not acquire the inputs for this planned production.

As for price/calculation issues: Firms also had incentives around cost (not necessarily to keep costs low, sometimes to inflate the paper-level of costs in order to argue for a higher price for themselves - but this is also a price-related issue) but the cost of inputs was not related to their actual value in terms of scarcity or consumer preference, hence when firms occasionally had some freedom to choose from among different kinds of inputs (what to make the product out of - the type of metal, cloth, etc) then they might choose an input that makes little sense for the final product, and reduces its quality considerably, based on this price. Just as you had 3x thicker sheets of glass and metal (from aiming to fill the output target faster) you also had many products made of bizarre materials that were actually chosen by factories because of the input price (or because of a lack of other options - coordination).

Anyway, in the Soviet Union these were all tied together. You could not simply say "they did not want to serve the customer, so they made the product badly" - although true, there was much more to it. Instead of serve the customer, firms were forced to fill targets, and faced with the input selection and relative prices that came from all other firms also having to fill targets.


Well, we're off the screen now, so this is probably almost a private conversation. Your points are valid, but I doubt all that important over the longer run. Partly because there was little technical change, there was good adjustment to current tech (see papers by White and others at end of 80s, 90 on tech efficiency of Soviet production, which was high). They did adjust to input problems. Now, it was slower, and certainly using inappropriate inputs was part of the problem. But there was also opportunity over time, assuming we are talking about a centrally planned good (not all were) to complain to higher authorities who, over time, would gradually adjust the quota requirements for the input producers.

The difference here is that firms in the centrally planned sector could get to GOSPLAN and sometimes get something done. Final consumers could not, so there was less feedback and incentive for any adjustment on final quality, although that could be affected by unresolved problems on input quality.

Barkley - it is true that complaints would eventually lead to adjustments, but there are two things to consider with regard to that kind of tattonement, if you like, in practice.

(1) The requirements may have changed somewhat by the time the new targets are set.

(2) Probably more important: in the Soviet Union targets were never precise about assortment (it was too difficult, given the limited knowledge about inputs & resources for the producer firm). Hence, the number of targets was increased and reduced again by several reforms, but they were never precise - they always left some part of the product mix up the firm. In addition, plan fulfillment relied not only on total output and assortment targets (fulfillment of the output mix), but also cost targets, wage fund targets, and myriad other targets. Hence, the firm could fulfill 90% of this target, fulfill that one 100% and another 110% and it would do well. For this reason, even if the assortment targets were correct, the firm was not all that likely to produce the appropriate amount of each of the products. But, the planners could not know which firms would produce the assortment completely and which would only produce to 90% (or which product the firm would produce to 90% or 100%) so the planners could not know which ones to set especially high--hence planners always planned with a "taut plan" trying to force firms to produce everything. This did not work, of course.

This is a summary of the "shortage economy" problem -- and it does explain a lot of coordination problems, and quality problems.

Basically, all I am arguing here is that you are highly discounting the true weight of the problems caused by planning--by prices, calculation, coordination, and there effects on quality. You seem to think incentives were the main problem, but they were only a minor part of the problem in truth. Things that you may assume were caused by a lack of incentives were actually caused by calculation and coordination problems.

their effects -

The comments to this entry are closed.

Our Books