Calculation in-Natura, from Neurath to

Kantorovich
Paul Cockshott

May 15, 2008

1 Introduction
Following the collapse of hither too existing socialism in Eastern Europe and
Central Asia. There was a crisis in socialist economic thought. If we contrast
the situation of the 1990s with what had existed 40 years earlier, we see that
whilst in the 1950s, socialism and economic planning were almost universally
accepted, even by enemies of socialism, as being viable ways to organize an
economy, by the 1990s the reverse applied. Among orthodox opinion it was
now taken for granted that socialism was the ’god that failed’, and that socialist
economic forms, when judged in the balance of history had been found want-
ing. And among socialist theorists there was a general retreat from ideas that
had previously been taken for granted, a movement towards market socialist
ideas, an accommodation with the idea that the market was a neutral economic
mechanism.
Whilst accommodation to the market was, to anyone familiar with Marx,
completely at odds with his critique of civil society[44], it nonetheless gained
considerable credence. Former governing socialist parties, thrown suddenly
into opposition in renascent capitalist states, felt that they had to restrict their
ambitions to reforms within a market economy.
In retrospect one can see that the mid 1970s represented the high water
mark of the socialist tide. Whilst the Vietnamese were driving the US out of
Saigon, and the last colonial empire in Africa, that of Portugal, was falling,
the collapse of the cultural revolution in China was setting the economic scene
for the triumph of capitalism in the 80s and 90s. When, after the death of
Mao, Deng threw open the Chinese economy to western capital investment,
the balance of economic forces across the whole world was upset. An im-
mense reserve army of labour, hireable of the lowest of wages, was thrown onto
the scales. The bargaining position of business in its struggles with domestic
labour movements was, in one country after another, immensely strengthened.
The general intellectual/ideological environment today is thus much less
favorable to socialism than it was in the 20th century. This is not merely a
consequence of the counter-revolutions that occurred at the end of the 20th
century, but stems from a new and more vigorous assertion of the classic tenets

1
of bourgeois political economy. This re-assertion of bourgeois political econ-
omy not only transformed economic policy in the West, but also prepared the
ideological ground for counter revolutions in the East.
The theoretical preparation for the turn to the free market that occurred in
the 1980s had been laid much earlier by right wing economic theorists like
Hayek and Friedman. Their ideas, seen as extreme during the 1950s and 60s
gained influence through the proselytizing activities of organizations like the
Institute for Economic Affairs and the Adam Smith Institute. These groups
produced a series of books and reports advocating free market solutions to con-
temporary economic problems. They won the ear of prominent politicians like
Margaret Thatcher, and from the 1980s were put into practice. She was given
the liberty to do this by a combination of long term demographic changes and
short term conjectural events. Within Britain, labour was in short supply, but
across Asia it had become super abundant. Were capital free to move abroad to
this plentiful supply of labour then the terms of the exchange between labour
and capital in the UK would be transformed. Labour would no longer hold the
stronger bargaining position. The conjunctural factor making this possible was
the surplus in foreign trade generated by North Sea oil. Hitherto, the workers
who produced manufactured exports had been essential to national economic
survival. With the money from the North Sea, the manufacturing sector could
be allowed to collapse without the fear of a balance of payments crisis.
The deliberate run-down of manufacturing industry shrank the social basis
of social democracy and weakened the voice of labour both economically and
politically.
The success of Thatcher in attacking the trades union movement in Britain
encouraged middle class aspiring politicians in the East like Vaclav Klaus and
presaged a situation in which Hayekian economic doctrines would become the
orthodoxy. Thatcher’s doctrine TINA, There Is No Alternative, (to capitalism)
was generally accepted.
The theoretical dominance of free market economic ideas had by the start
of the 21st century become so strong, that they were as much accepted by so-
cial democrats and self professed communists, as they had been by Thatcher.
They owe dominance both to class interests and to their internal coherence.
The capitalist historical project took as its founding documents the Declaration
of the Rights of Man, and Adam Smith’s Wealth of Nations. Together these
provided a coherent view of the future of Bourgeois or Civil Society, as a self
regulating system of free agents operating in the furtherance of their private
interests. Two centuries later when faced with the challenge of communism
and social democracy, the more farsighted representatives of the bourgeoisie
returned to their roots, restated the original Capitalist Manifesto, and applied it
to current conditions. The labour movement by contrast had no such coherent
social narrative. Keynes’s economics had addressed only technical issues of
government monetary and tax policy, it did not aspire to the moral and philo-
sophical coherence of Smith.
The external economic and demographic factors that originally favored the

2
turn to the market are gradually weakening. Within the next 20 years the vast
labour reserves of China will have been largely utilized, absorbed into capitalist
commodity production. Globally we are returning to the situation that West-
ern Europe had reached a century ago: a maturing world capitalist economy
in which labour is still highly exploited but is beginning to become a scarce
resource. These were the conditions that built the social cohesion of classi-
cal social democracy, the conditions that gave rise to the IWW and then CIO
in America, and led to the strength of communist parties in Western Europe
countries like France, Italy and Greece post 1945. We see perhaps, in South
America, this process in operation today.
These circumstances set 21st century critical political economy a new his-
torical project: to counter and critique the theories of market liberalism as
effectively as Marx critiqued the capitalist economists of his day.
The historical project of the world’s poor can only succeed if it promulgates
its own political economy, its own theory of the future of society. This new
political economy must be as morally coherent as that of Smith, must lead to
economically coherent policy proposals, which if enacted, open the way to a
new post-capitalist civilisation. As those of Smith opened the way to the post
feudal civilisation.
Critical political economy can no longer push to one side the details of
how the non-market economy of the future is to be organised. In the 19th
century this was permissible, not now. We can not pretend that the 20th century
never happened, or that it taught us nothing about socialism. In this task 20th
century Western critical Marxists like Cliff, Bettleheim or Bordiga will only
take us so far. Whilst they could point out weaknesses of hitherto existing
socialism, they did this by comparing it to an ideal standard of what these
writers thought that a socialist society should achieve. In retrospect we see
that these trends of thought were a product of the special circumstances of the
cold war, a striving for a position of ideological autonomy ’neither Moscow
nor Washington’, rather than a real contribution to political economy. The
very psychological detachment that such writers sought, deflecting from their
own heads the calumnies directed at the USSR, prevented them from positively
engaging with the problems faced by historically existing socialism. It is only
if you envisage being faced with such problems oneself, that one would come
up with practical answers:

"It is not the critic who counts: not the man who points out
how the strong man stumbles or where the doer of deeds could
have done better. The credit belongs to the man who is actu-
ally in the arena, whose face is marred by dust and sweat and
blood, who strives valiantly, who errs and comes up short again
and again, because there is no effort without error or shortcoming,
but who knows the great enthusiasms, the great devotions, who
spends himself for a worthy cause; who, at the best, knows, in the
end, the triumph of high achievement, and who, at the worst, if he

3
 fails, at least he fails while daring greatly, so that his place shall
never be with those cold and timid souls who knew neither victory
nor defeat." (Citizenship in a Republic, Roosevelt)

In the 19th century Marx’s Capital was a critique of the political economy that
underlay British Liberalism. 21st century critical political economy must per-
form an analogous critique of neo-liberal economics comparable in rigour and
moral depth. In particular it must engage with the ideas of the Austrian school:
Boehm-Bawerk, Mises, Hayek, whose ideas now constitute the keystone of
conservatism. Soviet Marxism felt strong enough to ignore them then, and the
response in the West came in the main from marginalist socialists like Lange
and Dickinson. If socialism is to reconstitute itself as the commonsense of the
21st century - as it was the commonsense of the mid 20th, then these are the
ideas that must be faced1 .
In attacking them one should not hesitate to use the advances in other sci-
ences - statistical mechanics, information theory, computability theory. And, to
re-establish scientific socialism there must be a definitive break with the spec-
ulative philosophical method of much of Western Marxism. From the time of
Marx till about the mid-twentieth century, most left intellectuals saw social-
ism and science as going hand-in-hand, in some sense. Most scientists were
not socialists (though some prominent ones were), but Marxists seemed to re-
gard science as friendly to, or consonant with, their project, and even saw it as
their duty as materialists to keep current with scientific thought and assess its
implications for social questions.
But since some point in the 1960s or thereabouts, many if not most West-
ern Marxist thinkers have maintained a skeptical or hostile attitude towards
science, and have drawn by preference on (old) philosophical traditions, in-
cluding Hegelianism. It is not clear why this has occurred but these may be
some of the factors:

• The conception of science as socially embedded. Science in bourgeois

society is bourgeois science, rather than offering privileged access to an
independent reality. This idea was obviously present in the Proletlult
tendency criticised by Lenin, and was later expressed in Lysenkoism.
In addition there has been a conflation of science and technology in the
minds of many writers. The role of nuclear weapons no doubt played
a part in this and spilled over to a general hostility to nuclear power.
Socio-biology too, was seen as hostile to progressive social thought, so
the alliance between Marxism and Darwinism came to be weakened.
Evolutionary psychology could be seen as transparent apologetics (for
example [47]), but this blinded left thinkers to progressive Darwinists
like Dawkins[15, 14].

• Althusser, the French communist philosopher was obviously pro-scientific

in intent, but may have unwittingly influenced many of his followers in
1 This article is part of a systematic program of work aimed at contributing to this critique,

previous articles were [10, 5].

4
 a contrary direction. One could easily get the impression from Althusser
that while staying too close to Hegel is an error, empiricism is a cardinal
sin. Equate empiricism and science, and you’re off to the races.

• The appropriation of the "Scientific Socialism" label by the USSR and

its official ideologists.

• The brute historical fact that while science was doing very nicely, social-
ism in the West was not. Thus undermining the idea that Marxism and
science somehow marched together.

Whatever exactly is the cause, the effect is that while in the 1930s (say) one
might have expected the "typical" young Marxist intellectual to have a scien-
tific training – or at least to have general respect for scientific method – by
the turn of the century one would be hard pressed to find a young Marxist in-
tellectual (in the dominant Western countries) whose background was not in
sociology, accountancy, continental philosophy, or perhaps some "soft" (quasi-
philosophical) form of economics, and who was not profoundly skeptical of
(while also ignorant of) current science2 .
Unlike that Western Marxist tradition have to treat political economy and
the theory of social revolution like any other science. We must formulate
testable hypotheses, which we then asses against empirical data. Where the
empirical results differ from what we expected, we must modify and retest our
theories3 .
In addition we must recover and celebrate the advances in political econ-
omy that arose from the Russian experience: the method of material balances
used in preparing the 5 year plans and systematized as Input Output analysis
by Leontief; the method of linear programming pioneered by Kantorovich; the
time diaries of Strumlin.
In this article I focus particularly on recovering the work of Kantorovich the
only Soviet Economic Nobelist, and showing that his work provided a funda-
mental theoretical response to von Mises. Kantorovich was an eminent mathe-
matician, whose work went well beyond economics, but in this article I focus
only on his economic contributions. In explaining these I reproduce in section
3.3 some of his original numerical examples drawn from his experience in So-
viet heavy industry. I have avoided, however, giving any detailed presentation
of the mathematical techniques ( algorithms ) that Kantorovich and Dantzig de-
veloped, both because I assume that the readership are not specialists in linear
algebra, and secondly because these techniques have now been packaged up
in open-source software that can be used by non-algebraists. I give in sections
3.3 and 3.4 what is essentially a tutorial introduction to using such package
to solve planning problems. I summarise what these mathematical techniques
mean in practical terms. What type of economic problem do they allow us to
solve?
2
I owe the above argument about Western Marxism to my co-worker Allin Cottrell.
3 For work in the vein see [6, 38, 7, 63, 64, 51, 8].

5
 As illustrations I will focus on how Kantorovich allows us to pose problems
of national or continental environmental trade-offs. From this I go on to ask
how do his ideas relate to the Austrian critique of socialism?
What are their implications for the future of economic planning?
How has the field advanced since Kantorovich’s day, and what are the po-
litical implications of these advances?

2 What is economic calculation?

In contemporary society the answer seems simple enough: economic calcula-
tion involves adding up costs in terms of money. By comparing money costs
with money benefits one may arrive at a rational - wealth maximizing - course
of action.
In a famous paper[58] the Austrian economist Mises argued that it was only
in a market economy in which money and money prices existed, that this sort
of economic rationality was possible.
His claims were striking, and, if they could be sustained, apparently devas-
tating to the cause of socialism. The dominant Marxian conception of social-
ism involved the abolition of private property in the means of production and
the abolition of money, but Mises argued that "every step that takes us away
from private ownership of the means of production and the use of money also
takes us away from rational economics" ([58]: 104). The planned economy
of Marx and Engels would inevitably find itself "groping in the dark", produc-
ing "the absurd output of a senseless apparatus" (106). Marxists had counter-
posed rational planning to the alleged ‘anarchy’ of the market, but according to
Mises such claims were wholly baseless; rather, the abolition of market rela-
tions would destroy the only adequate basis for economic calculation, namely
market prices. However well-meaning the socialist planners might be, they
would simply lack any basis for taking sensible economic decisions: socialism
was nothing other than the "abolition of rational economy".
As regards the nature of economic rationality, it is clear that Mises has in
mind the problem of producing the maximum possible useful effect (satisfac-
tion of wants) on the basis of a given set of economic resources. Alternatively,
the problem may be stated in terms of its dual: how to choose the most efficient
method of production in order to minimize the cost of producing a given use-
ful effect. Mises repeatedly returns to the latter formulation in his critique of
socialism, with the examples of building a railway or building a house:4 how
can the socialist planners calculate the least-cost method of achieving these
objects?
As regards the means for rational decision-making, Mises identifies three
possible candidates: planning in kind (in natura ), planning with the aid of
an ‘objectively recognizable unit of value’ independent of market prices and
money, such as labour time, and economic calculation based on market prices.
4 The railway example is in [58]. The house-building example is in Human Action [60].

6
I will go on to examine Mises’, very influential, arguments in section 3, but
first I will examine whether an alternative interpretation can be placed on the
concept of economic calculation.
It is clear that monetary calculation lends itself well to problems of the
minimising or maximising sort. We can use money to find out which of several
alternatives is cheaper, or which sale will yield us the most profit. But if we
look in more detail at what is involved here, we shall see that a lot of calculation
must take place prior to money even being brought into consideration. Let us
look not at building a mere house, but at something grander, the first Pyramid
at Saqqara, planned by Imhotep[4]. In order to build this Imhotep had to carry
out a whole mass of calculations. He needed, for example, to know how to
calculate the volume of pyramid before it was built([32], p.40 ), which involves
a fair degree of sophisticated geometry5. From a knowledge of the volume of a
pyramid, and a knowledge of the size of the stones he planned to use, he could
calculate how many stones would be required. Knowing the rate at which
stonemasons could put the stones in place he could estimate how long it would
take workforces of different sizes to place all the stones for the pyramid. From
the number of stones too, and knowledge of how many people are needed to
transport each stone, Imhotep could work out the number of people who would
have to work shifting the stone from the quarry to the pyramid.
This workforce would have to be fed, so bakers, brewers and butchers were
needed to feed them ([13],ch.6). He, or his scribes, would have to calculate
how many of these tradesmen were required. Quantities of grain and cattle
would have to to be estimated. In the broadest sense, this was all economic
calculation, but it would have taken place without money, which had yet to be
invented. It might be objected that this is not what Mises meant by economic
calculation, since Imhotep’s calculation ’in kind’ was not economic calculation
but engineering calculation, a mere listing of prerequisites, what was missing
was the valuation or costing of these inputs. Fair enough, this is not what von
Mises meant by economic calculation, the question is, whether he was right
to limit this concept to monetary calculation. Imhotep’s calculations do reveal
that Mises concept may have been too narrow. Suppose that the pyramid were
built now, a large part of the calculations required would be the same. It would
still be necessary to work out how much stone would be used, how much of
various types of labour would be used, how the stone was to be transported
etc. This would be the difficult part of the calculation, totaling it up in money
would be easy in comparison.
Consider the issue of choosing between the most economical alternative.
Imhotep certainly had to address this question. Building a pyramid was, even
by modern standards, a massive undertaking. To complete it he not only had
to address questions of structural stability but he also had to devise a practical
method by which stones could be raised into place. That this was no easy task

5 The Rhind Papyrus, the earliest known collection of mathematical problems, includes exam-

ples where the student had to calculate the volume of, and thus the number of bricks required for,
pyramids.

7
is born out by the fact that we still do not know for sure how it was done. Vari-
ous suggestions have been made: sloping ramps at right angles to the pyramid
wall up which stones were hauled; spiral ramps wrapping round the pyramid;
internal tunnel ramps; a series of manually operated cranes; etc. If we today
can think of lots of possible ways in which it might be done, so to, we can as-
sume, must the original builders, before settling on whatever method that they
actually used. The resources of manpower available to them were not unlim-
ited, so they had to discover an approach that was both technically feasible and
economically feasible. This is the sort of rational choice that Mises saw as
impossible without money, but the fact that the pyramids were built, indicates
that some calculation of this type did occur.
The ultimate constraint here was the labour supply available; no sensible
architect would embark on a course of construction that used far more labour
than another. In a pre-mercantile economy like ancient Egypt this labour con-
straint appears directly, in a mercantile economy, the labour constraint appears
indirectly in the form of monetary cost. The classical political economists
argued that money relations disguised underlying relations of labour, money
costs hid labour costs; money was, for Adam Smith, ultimately the power to
command the labour of others.

3 Planning in kind
The organisational task that faced a pyramid architect was vast. That it was
possible without money was an indication that monetary calculation was not a
sine qua non of calculation. But as the project being planned becomes more
complex, then planning it in material units will become more complex. Mises
is in effect arguing that optimization in complex systems necessarily involves
arithmetic, in the form of the explicit maximization of a scalar objective func-
tion (profit under capitalism being the paradigmatic case), and that maximising
the money return on output, or minimising money cost of inputs is the only
possible such scalar objective function. Mises argued for the impossibility of
of planning in kind because, he said,. the human mind is limited in the degree
of complexity that it can handle.
So might the employment of means other than a human mind make possible
planning in kind for complex systems?
There are two ’inhuman’ systems to consider:

1. Bureaucracies. A bureaucracy is made up of individual humans, but by

collaborating on information processing tasks, they can carry out tasks
that are impossible to one individual.

2. Computer networks. Nobody familiar with the power of Google6 to con-

solidate and analyse information will need persuading that computers
6 The algorithms used by Google involve the solution to large sparse systems of linear equations.

This, as we shall see later, is the same type of calculation as is required for planning in kind. For
a discussion of the linear algebra used in information retrieval see the book by Google researcher
Dominic Widdows[62] or [57].

8
 can handle volumes and complexities of information that would stupefy
a single human mind, so a computer network could clearly do economic
calculations far beyond an individual human mind.

More generally as Turing pointed out [55] any extensive calculation by human
beings depends on artificial aides-memoir, papyrus, clay tablets, slates, etc.
With the existence of such aides to memory, algorithmic calculation becomes
possible, and at this point the difference between what can be calculated by a
human using paper and pencil methods or a digital computer come down only
to matters of speed[53, 54]. There is thus no difference in principle between
planning using a bureaucracy and planning using computers, but there is in
practice a big difference in the complexity of problem that can be expeditiously
handled.
There is no question that the procedure of economic calculation considered
by von Mises was primarily algorithmic. It involves a fixed process of

1. For each possible technique of production

(a) form a physical bill of materials,

(b) use a price list to convert this into a list of money expenditures,
(c) then add up the list to form a final cost

2. Select the cheapest final cost out of all the costs of techniques of produc-
tion

We will come back to Mises’s problem after looking at the views of his oppo-
nent Neurath.

3.1 Neurath’s original argument.

Mises was initially debating against Otto Neurath. In an article dated 1919
Neurath had argued that a socialist economy would be able to operate calcula-
tions in-natura rather than by means of money[42], though he arguably did not
provide a practical means of doing this [56]. Mises is much better known in
the English speaking world than Neurath, in large part because translations of
Neurath’s economics works have only recently appeared. The fact that Mises’s
readers have not had direct access to the ideas against which Mises was ar-
guing may have helped the plausibility of Mises’s argument. It is thus worth
recapitulating what Neurath meant by calculation in kind, so that one can asses
to what extent Mises’s criticisms were fair.
In his 1919 paper, Neurath argues that the experience of the war economy
allowed one to see certain key weaknesses of past economic thought.

Conventional economic theory mostly stands in too rigid a con-

nection to monetary economics and has until now almost entirely
neglected the in-kind economy.([42], p 300)

The war economy had in contrast been largely an in-kind economy.

9
 As a result of the war the in-kind calculus was applied more often
and more systematically than before... It was all to apparent that
war was fought with ammunition and the supply of food, not with
money.( [42], p304)

In kind views of quality of life. He argues that this represents a return to

the original concerns of economics in the science of household economics and
the science of government. Smith had been particularly concerned with the
real rather than the monetary income of society, but this had been forgotten by
subsequent economists who had concentrated on monetary magnitudes. Neu-
rath advocated an explicitly Epicurean approach to economics identifying his
approach as social Epicureanism. Neurath claimed that this Epicureanism lay
at the basis of Marx’s thought too, though if Marx’s doctoral dissertation is to
be believed[34], Neurath’s emphasis on the empirical investigation of real con-
ditions owes more to Democritus. If one wanted to know whether real quality
of life of the population was improving or not one had to examine their lives in
material not money terms[41]. He wrote that economics must be the study of
happiness and the quality of real life. To do this economists should collect de-
tailed statistics of the quality of life of groups in the population. These would
include not only on the consumption of food, clothing and housing conditions,
but also on mortality and morbidity, educational level, leisure activities, peo-
ple’s feelings of powerfulness versus powerlessness.

With some expectation of success we can attempt to assemble

all conditions of life into certain larger groups and arrange them
according to the pleasurableness of the qualities of life caused by
them. We can, for example, state what food the individuals con-
sume per year, what their housing conditions are, what and how
much they read. what their experiences are in family life, how
much they work, how often and how seriously they fall ill, how
much time they spend walking, attending religious services, enjoy-
ing art, etc. We can even discover certain average biographies, de-
viations from which appear unimportant for rough investigations.
In similar ways we can also determine the conditions of life of
whole groups of people by stating which proportion of them suf-
fer from certain ailments, which proportion dies at a certain age,
which proportion lives in certain homes. etc., finally even which
proportion enjoys particular types of conditions of life. It is obvi-
ous that quantities which can be measured and determined clearly
find more extensive treatment than the vaguer ones like religiosity,
artistic activities and the like. But one must beware of thinking
that all those quantities which can be treated more easily are more
important, or essentially different from the vague ones. Occupa-
tional prestige, for example, is as much a part of one’s income as
eating and drinking. ([41] page 326)

10
Compared to such statistics in kind, figures for national income were, he said,
far less revealing. In particular he cautions against accepting the notion of
’real income’ or inflation adjusted money income as a surrogate for the quality
of life. Such ’real income’ is just a reflection of money income and as such
only takes into account things that are bought and sold as commodities.

The current concept of consumption, [so-called] real income,

is also understandable as derivative of money calculation. Given
our own approach to economic efficiency, it seems appropriate to
comprehend also :work and illness under the concept which covers
food, clothing, housing, theatre visits, etc. These things, however,
are not part of the [current] concept of consumption and real in-
come, which covers only what appears as a reflection of money
income. Real income [in this sense] has little significance in our
approach to the study of economic efficiency. ( [41] page 336)

What Neurath was saying here looks very modern. There has been increasing
recognition of the inadequacy of purely monetary national income figures for
judging the quality of life of a country’s population (sources???). The UN
development goals are informed by such concerns and are given in qualitative
terms.... (cite). It is notable that this aspect of Neurath’s argument for in-
kind economics has been neglected by von Mises or his followers. Indeed
Neurath argues that von Mises himself ultimately has recourse to the notion of
an in-kind substratum of welfare against which different monetary measures of
welfare must be judged. Mises recognises that monopoly reduces welfare thus:

He (Mises) arrives at the remarkable statement: "But these,

of course, are less important goods, which would not have been
produced and consumed if the more pressing demands for a larger
quantity of the monopolized commodity could have been satisfied.
The difference between the values of these goods and the higher
value of the quantity of monopoly goods not produced represents
the loss in welfare which the monopoly has inflicted on the na-
tional economy."7 We see that here Mises also arrives at a concept
of wealth which obviously is divorced from money, since it is used
to assess a money calculation, namely that of the monopolists. If,
in the case of monopoly, according to Mises, there is a calcula-
tion of wealth by which one can judge money calculation, then it
should always be available and allow judgment on all economic
processes. ([40], page 429)

Neurath is here defending the distinction between exchange value and use value
which comes from Aristotle[2, 36] and provided a key substratum of Marx’s
analysis of the commodity[35].
7 Neurath is citing [59]page 389.

11
In kind calculation for production Neurath was adamant that a socialist
economy had to be moneyless. In this, he was an orthodox follower of Marx,
and as such much more radical than the Soviet government post war-communism.
He repeatedly emphasizes that a socialist economy can not use just one single
scalar unit in its calculations, whether this be money, labour hours or kilowatt
hours. This relates both to :

1. The non-comensurability of final outcomes in terms not only of quality

of life, but the quality of life of future generations;

2. The complexity of the technical constraints on production.

The emphasis on non-comensurability has its roots in his ideas on the measure-
ment of outcomes, quality of life now and quality of life in the future:

The ’positive quantities’ of the socialist order also do not come

to the same thing as the ’profit’ of capitalism. Savings in coal,
trees, etc., beyond amounting to savings in the displeasure of work,
mean the preservation of future pleasure, a positive quantity. For
instance, that coal is used nowadays for silly things is to be blamed
for people freezing in the future. Still, one can only give vague es-
timates. Saving certain raw materials can become pointless if one
discovers something new. The future figures in the balance sheets
of the capitalist order only in so far as the demand is anticipated.
The freezing people of the future only show up if there is already
now a demand for future coal.([40], page 470)

Neurath follows Marx in accepting the use of labour vouchers as a possible

means of distributing goods, provided that the community decides to do it this
way, but denies that this method has anything more than a conventional sig-
nificance. In particular he argues that labour time calculations are inadequate
for the internal regulation of production. Labour time calculations presuppose
a long time frame and an absence of natural resource constraints. If there are
natural resource constraints, or short term shortages of particular equipment
they can misrepresent what is potentially producible.

How can points be assigned to individual articles of consump-

tion? If there were natural work units and if it could be determined
how many natural work units, in a “socially necessary” way, have
been spent on each article of consumption, and if further it were
possible to produce any amount of each article, then, under some
additional conditions, each article could be assigned the number
of points that represent its “work effort”. [. . . ] Let us now assume
that the distribution is done through free choice of the consumers
in proportion to their work. [. . . ] some raw materials will be in
short supply and thrift will necessary. If there is a great demand
for articles made from these raw materials, either rationing will

12
 have to be introduced or the number of points for their distribu-
tion will have to be increased beyond the number representing the
work spent on their production. Conversely articles in little de-
mand will be offered for fewer points than would the work spent
for their production. ( [40], pp. 435-436)

These do not seem to be insuperable obstacles to the use of labour vouchers in

distribution of final products. One could conceive of there being some sort of
natural resource tax levied on goods whose production could not be expanded
until the number of points for their distribution was equal to the work spent
on their production. The proceeds of this tax could then contribute towards
the labour expended providing free public services. But the point about labour
values being insufficient for the internal regulation of production is correct.
Instead he advocates detailed statistics on the consumption and use of each
raw material and intermediate product. He proposes a system with two tables
in kind for each raw material and intermediate product X

1. One table gives, in quantitative terms the output of X product, the im-
ports and exports, and all the uses. He gives an illustration in which he
shows the flows stocks and use of copper ore in Germany between 1918
and 1919.

2. Another table gives for X all the raw materials, types of labour and in-
termediate products that went into making it.

Accounting balances in kind will be used to check the correctness of the pro-
duction and uses between these different tables. If we look at this we can see
that although he presents this in terms of distinct tables, these tables record the
same information as respectively to the row and column vectors of an input
output matrix. The one key difference is that current western I/O matrices list
all quantities in the matrix in money, whereas Neurath proposes listing them in
natural units : tons, litres etc. Since the work of von Neumann ( discussed be-
low ) we have become used to representing the technical structure, the in-kind
flows, of the economy in matrix form. By using matrices it becomes possible to
express propositions about the economy in the concise notations of the matrix
and vector algebra, and to have recourse to the theorems of that algebra. But
there is a big difference between constructing abstract mathematical proofs and
carrying out practical economic administration.
The matrix notation of von Neumann is certainly more elegant in math-
ematical terms, but, as a practical tool for economic calculation, Neurath’s
system has great advantages. Suppose that in Germany in 1919 there were
200,000 distinct industrial products to be tracked. We know from current I/O
tables that one can print a table of perhaps 80 products square on an A3 page.
The complete von Neumann or Leontief style I/O matrix for 200,000 products
would then run to over 6 million pages. The great bulk of these entries would
be blank. To take Neurath’s example of copper ore, there might be a couple
of dozen copper foundries using the ore, so the copper ore row of a complete

13
von Neumann I/O matrix, would run blank ( or zeros) for thousands of pages.
Neurath’s usage table for copper ore, could on the other hand, be printed on
a single page. The representation advocated by Neurath is actually similar to
that used in modern computing when dealing with large matrices, where it is
called a ’sparse matrix’ representation. The advantages of this representation
for computerized planning are examined in [9] chapter 6.
But if we stick for a moment with the matrix notation familiar to modern
economists, we can understand why Neurath was so adamant that socialist cal-
culation had to be performed in kind and could not be reduced to accounting in
a single surrogate unit like labour or energy. When we do accounting in money,
or in a surrogate like labour, then we add up the total cost of each column of the
I/O matrix, giving us a vector of final output in money terms. A price system
thus represents an enormous destruction of information. A matrix of technical
coefficients is folded down to a vector, and in the process the real in-natura
constraints on the economy are lost sight of. This destruction of information
means that an economy that works only on the basis of the price vector must
blunder around with only the most approximate grasp of reality. This of course,
is exactly the opposite proposition to that advanced by Mises.
To summarise, Neurath had argued that in kind calculation was needed both
to allow political deliberation on the goals of the economic plan, and to ensure
the coherence of the plan. Mises has no effective reply to the first point, and
concentrated his fire on the second. Mises concedes that if there is no change
in technique then the sort of in-kind accounting proposed by Neurath would
allow the continued operation of the socialist economy. The problem came
in choosing between competing techniques. Whilst Neurath clearly believed
that this was possible, he is vague about how it is to be done. He does not
give a procedure or algorithm by which assessments of comparative technical
efficiency can be arrived at using in-kind calculation.
The question then arises as to whether, independent of the work of Neurath,
there exist in-natura algorithms with a function analogous to those that Mises
saw as essential for economic calculation?
We will argue that subsequent authors, working in the two decades after
Neurath’s proposals, did in fact come up first with mathematical proofs that
there exist solutions to a system of calculation in kind, and then with practical
algorithms to arrive at such solutions.

3.2 von Neumann

The next two players in our drama have certain similarities. Both von Neu-
mann and Kantorovich were mathematicians rather than economists. Their
contributions to economics were just one part of a variety of research achieve-
ments. In both cases this included stints working on early nuclear weapons
programs, for the US and USSR[50] respectively. At least in von Neumann’s
case the connection of his economic work to atomic physics was more than
incidental. One of his great achievements was his mathematical formalization

14
of quantum mechanics[61] which unified the matrix mechanics of Heisenberg
with the wave mechanics of Schrodinger. His work on quantum mechanics co-
incided with the first draft of his economic growth model[39] given as a lecture
in Princeton in 1932. In both fields he employs vector spaces and matrix opera-
tors over vector spaces, complex vector spaces in the quantum mechanical case,
and real vector spaces in the growth model. Kurz and Salvadori [30]argue that
his growth model has to be seen as a response to the prior work of the socialist
inclined mathematician Remak[48], who worked on ’superposed prices’.

Remak then constructs ‘superposed prices’ for an economic

system in stationary conditions in which there are as many single-
product processes of production as there are products, and each
process or product is represented by a different ‘person’ or rather
activity or industry. The amounts of the different commodities ac-
quired by a person over a certain period of time in exchange for his
or her own product are of course the amounts needed as means of
production to produce this product and the amounts of consump-
tion goods in support of the person (and his or her family), given
the levels of sustenance. With an appropriate choice of units, the
resulting system of ‘superposed prices’ can be written as

pT = pT C

where C is the augmented matrix of inputs per unit of output, and

p is the vector of exchange ratios. Discussing system Remak ar-
rived at the conclusion that there is a solution to it, which is semi-
positive and unique except for a scale factor. The system refers
to a kind of ideal economy with independent producers, no wage
labour and hence no profits. However, in Remak’s view it can also
be interpreted as a socialist economic system.[30]

With Remak the mathematical links to the then emerging matrix mechanics are
striking - the language of superposition, the use of a unitary matrix operator C
analogous to the Hermitian operators in quantum mechanics8. But this apart,
what is the economic significance of Remak’s theory to the socialist calculation
debate?
It is this. Remak shows for the first time how, starting from an in-natura
description of the conditions of production, one can derive an equilibrium sys-
tem of prices. This implies that the in-natura system contains the information
necessary for the prices and that the prices are a projection of the in-natura
system onto a lower dimensional space9 . If that is the case, then any calcu-
lations that can be done with the information in the reduced system p could
8 Like the Hermitian operators in quantum mechanics, Remak’s production operator is unitary

because pis an eigen vector of C and |p| is unchanged under the operation.
9 Suppose C is an n × n square matrix, and p an n dimensional vector. By applying Iverson’s

reshaping[24, 23] operator ρ , we can map C to a vector of length n2 thus c ← (n × n)ρ C , and we
thus see that the price system, having ndimensions involves a massive dimension reduction from
the n2 dimensional vector c.

15
in principle be done, by some other algorithmic procedure starting from C.
Remak expresses confidence that with the development of electric calculating
machines, the required large systems of linear equations will be solvable.
The weakness of Remak’s analysis is that it is limited to an economy in
steady state. Mises had acknowledged that socialist calculation would be pos-
sible under such circumstances.
Von Neumann took the debate on in two distinct ways:

1. He models an economy in growth, not a static economy. He assumes

an economy in uniform proportionate growth. He explicitly abjures con-
sidering the effects of restricted natural resources or labour supply, as-
suming instead that the labour supply can be extended to accommodate
growth. This is perhaps not unrealistic as a picture of an economy un-
dergoing rapid industrialization ( for instance Soviet Russia at the time
he was writing ).

2. He allows for there to be multiple techniques to produce any given good -

Remak only allowed one. These different possible productive techniques
use different mixtures of inputs, and only some of them will be viable.

von Neumann again uses the idea of a technology matrix introduced by Remak,
but now splits it into two matrices A which represents the goods consumed in
production, and B which represents the goods produced. So ai j is the amount
of the j th product used in production process i, and bi j the amount of product
j produced in process i. This formulation allows for joint production, and
he says that the depreciation of capital goods can be modeled in this way, a
production process uses up new machines and produces as a side effect older,
worn machines. The number of processes does not need to equal the number of
distinct product types, so we are not necessarily dealing with square matrices.
Like Remak he assumes that there exists a price vector y but also an in-
tensity vector x which measures the intensity with which any given production
process is operated. We will see below that the same formulation is used by
Kantorovich. Two remaining variables β and α measure the interest rate and
the rate of growth of the economy respectively.
He makes two additional assumptions. First is that there are ’no profits’,
by which he means that all production processes with positive intensity return
exactly the rate of interest. He only counts as profit, earning a return above the
rate of interest. This also means that no processes are run at a loss ( returning
less than β ). His second assumption is that any product produced in excessive
quantity has a zero price.
He goes on to show that in this system there is an equilibrium state in which
there is a unique growth rate α = β and definite set of intensities and prices.
The intensities and prices are simultaneously determined.
What are the significant results here?

• The in − natura techniques available to the economy, captured in his

matrices A, B determine which processes of production should be used

16
 and in which intensities.

• They also determine an equilibrium set of prices. No system of subjec-

tive preferences is required to derive these.

• The in-natura techniques also determine the rate of growth and rate of
interest.

What are the social relations in this model?

It is unclear. If it is a capitalist economy he is making the rather unrealistic
assumption that all interest income is reinvested, so that interest becomes not
so much a payment to the bank as an accounting convention. It is also unclear
how a real capitalist economy could reach the equilibrium path shown. Sraffa
[52] presents a rather similar model, explicitly identifying it with capitalist
production, but with the crucial addition that Sraffa allows for the possibil-
ity of capitalist consumption out of interest. In the absence of any capitalist
consumption, the interpretation of von Neumann’s model as being of an ad-
ministrative economy, is plausible. However, it is an administrative economy
with at least accounting prices and a notional accounting charge for capital use.
If he means that the economy is to be understood as capitalist, then he should
really prove that his twin conditions of zero prices for goods in excess supply
and an absolutely uniform rate of profit, can be achieved by market compe-
tition. Showing this would have been non-trivial. Indeed there is reason to
suspect that uniform profit rates can not be achieved in dynamic models of this
type[18].
If we suppose that von Neumann is describing an administrative economy,
then it is significantly different from Neurath’s idea, because of the existence
of at least an administrative price vector. But this price vector is shown to arise,
along with the interest rate, purely from the in-kind structure of the economy,
so, as with Remak, prices are a derived sub-space. Von Neumann’s paper does
not, however, provide a procedure by which the equilibrium solution to the
economy can be calculated. He proves the existence of such a solution but
does not give a means of computing it.
If we have no joint production and only one process to produce each prod-
uct, it is relatively simple to solve the VN model. Suppose we have several
product types one of which is corn, with the von Neumann matrices A, B such
that both are square and B = I. Suppose further that we have the variables
in table 1, then Algorithm 1 will find the prices, growth rate, and intensities
arbitrarily close to the von Neumann solution depending on ε .
If A, B take on the values given in Table 2, then with ε = 0.001the algorithm
gives the approximate solution shown in the lower part of Table 2.

3.3 Kantorovich’s method

In the early 30s, no algorithmic techniques were known which would solve
the more general problem where there can be joint production and multiple
possible techniques to produce individual products. But in 1939 [25] the Soviet

17
 Table 1: Variables used in algorithm 1.
variable meaning
x intensity vector
n net output vector
µ inputs used
y price vector denominated in corn
c per unit cost vector in corn
β interest rate
α growth rate
sales total sales in corn units
costs total costs in corn units

Algorithm 1 Solving a VN model with no choice of techniques.

begin
initial intensities x← T;
initial pirces y← 1;
estimated interest β ← 0.2;
repeat
α← β ;
compute cost per unit c ← (A.y ) × (1 + β );
set prices y ← c;
y corn ← 1;
compute usage µ ← ∑ (( A T ) × x ) ;
sales ← x.y ;
n← x - µ ;

costs ← y. µ ;

recompute interest β ← sales−costs

costs
;
x ← 0.5 × (x + µ × (1 + α ));

the above line will make y move towards a composition in which the physical
proportions of inputs and outputs are the same

until |β − α | < ε ;

end .

18
 Table 2: Example A and B matrices and the VN solution they give rise to.

A
corn coal iron
0.20 0.10 0.02
0.20 0.20 0.10
0.20 0.70 0.10
B
1.00 0.00 0.00
0.00 1.00 0.00
0.00 0.00 1.00

Solution

corn coal iron

n 3.11427 3.46149 1.02518
y 1.00000 1.80357 3.56645
x 6.09637 6.88489 2.04303
β = 1.01806
α = 1.01866

mathematician V Kantorovich came up with a method which later came to be

known as linear programming or linear optimisation, for which he was later
awarded both Stalin and Nobel prizes. Describing his discovery he wrote:

I discovered that a whole range of problems of the most di-

verse character relating to the scientific organization of production
(questions of the optimum distribution of the work of machines
and mechanisms, the minimization of scrap, the best utilization of
raw materials and local materials, fuel, transportation, and so on)
lead to the formulation of a single group of mathematical problems
(extremal problems). These problems are not directly comparable
to problems considered in mathematical analysis. It is more cor-
rect to say that they are formally similar, and even turn out to be
formally very simple, but the process of solving them with which
one is faced [i.e., by mathematical analysis] is practically com-
pletely unusable, since it requires the solution of tens of thousands
or even millions of systems of equations for completion.
I have succeeded in finding a comparatively simple general
method of solving this group of problems which is applicable to
all the problems I have mentioned, and is sufficiently simple and
effective for their solution to be made completely achievable under
practical conditions. ([25], p. 368)

What was significant about Kantorovich’s work was that he showed that it was
possible, starting out from a description in purely physical terms of the various
production techniques available, to use a determinate mathematical procedure
to determine which combination of techniques will best meet plan targets. He

19
 Table 3: Kantorovich’s first example.
Type of machine # machines output per machine Total output
As Bs As Bs
Milling machines 3 10 20 30 60
Turret lathes 3 20 30 60 90
Automatic turret lathes 1 30 80 30 80
Max total 120 230

Table 4: Kantorovich’s examples of output assignments.

Type of machine Simple solution Best solution
As Bs As Bs
Milling machines 20 20 26 6
Turret lathes 36 36 60 0
Automatic turret lathes 21 21 0 80
Total 77 77 86 86

indirectly challenged von Mises10 , both by proving that in-natura calculation

is possible, and by showing that there can be a non monetary scalar objective
function : the degree to which plan targets are met.
The practical problems with which he was concerned came up whilst work-
ing in the plywood industry. He wanted to determine the most effective way of
utilising a set of machines to maximise output. Suppose we are making a final
product that requires two components, an A and a B. Altogether these must
be supplied in equal numbers. We also have three types of machines whose
productivities are shown in the Table 3.
Suppose we set each machine to produce equal numbers of As and Bs. The
three milling machines can produce 30 As per hour or 60 Bs per hour. If the
three machines produce As for 40 mins in the hour and Bs for 20 mins then they
can produce 20 of each. Applying similar divisions of time we can produce 36
As and Bs on the Turret lathes and 21 As and Bs on the automatic turret lathe
(Table 4).
But Kantorovich goes on to show that this assignment of machines is not
the best. If we assign the automatic lathe to producing only Bs, the turret lathe
to producing only As and split the time of the milling machines so that they
spend 6 mins per hour producing Bs and the rest producing As, the total output
per hour rises from 77 As and Bs to 86 As and Bs.
The key concept here is that each machine should be preferentially assigned
to producing the part for which it is relatively most efficient. The relative
efficiency of producing As/Bs of the three machines was milling machine = 12 ,
2
turret lathes = 3 , and automatic lathe = 83 . Clearly the turret lathe is relatively
most efficient at producing As, the automatic lathe is relatively most efficient at
producing Bs and the milling machine stands in between. Thus the automatic
lathe is set to produce only Bs, the turret lathes to make only As and the time
of the milling machines is split so as to ensure that an equal number of each
10 There is no indication that he was aware of von Mises at the time.

20
 auto turret lathe Plan Ray

A
milling machine

turret lathe

Figure 1: Kantorovich’s example as a diagram. The plan ray is the locus all
points where the output of As equals the output of Bs. The production possibil-
ity frontier is made of straight line segments whose slopes represent the relative
productivities of the various machines for the two products. As a whole these
make a polygon. The plan objective is best met where the plan ray intersects
the boundary of this polygon.

product is turned out.

The decision process is shown diagrammatically in Figure 1. The key to
the construction of the diagram, and to the decision algorithm is to rank the
machines in order of their relative productivities. If one does this, one obtains
a convex polygon whose line segments represent the different machines. The
slopes of the line segments are the relative productivities of the machines. One
starts out on the left with the machine that is relatively best at producing Bs,
then move through the machines in descending order of relative productivity.
Because relative productivity is monotonically decreasing one is guaranteed
that the boundary will be convex. One then computes the intersection of the 45
degree line representing equal output of As and Bs with the boundary of this
polygon. This intersection point is the optimal way of meeting the plan. The
term linear programming stems from the fact that the production functions are
represented by straight lines in the case of 2 products, planes for 3 products,
and for the general higher dimensional case by linear functions. That is to say,
functions in which variables only appear raised to the power 1.
The slope of the boundary where the plan ray intersects was called by Kan-
torovich the resolving ratio. Any machine whose slope is less than this should
be assigned to produce Bs any machine whose slope is greater, should be as-
signed to produce As.
When there are only two products being considered, the method is easy
and lends itself to diagrammatic representation. But it can handle problems of
higher dimensions, involving 3 or more products. In these cases we can not
use graphical solutions, but Kantorovich provided an algorithmic by which the
resolving ratios for different pairs of outputs could be arrived at by succes-
sive approximations. Kantorovich’s work was unknown outside of the USSR
until the late 50s and prior to that Dantzig had independently developed a sim-

21
Algorithm 2 Kantorovich’s example as equations input to lp_solve..
A;
m1<=3;
m2<=3;
m3<=1;
A-B=0;
m1-0.1 x1a - 0.05 x1b=0;
m2-0.05 x2a - 0.033333 x2b=0;
m3- 0.033333 x3a - 0.0125 x3b=0;
x1a+x2a+x3a - A=0;
x1b+x2b+x3b -B =0;
int A;

ilar algorithm for solving linear programming problems, the so called simplex
method [12]. This has subsequently been incorporated into freely available
software tools11 . These packages allow you to enter the problem as a set of
linear equations or linear inequalities which they then solve.
In the West, linear programming was used to optimise the use of production
facilities operating within a capitalist market. This meant that the objective
function that was maximised was not a fixed mix of outputs, in Kantorovich’s
first example equal numbers of parts A and B, but the money that would be
obtained from selling the output: price A ×number of As + price B ×number
of Bs. Manuals and textbooks produced in association with Western linear
programming software assumes this sort of objective. However, as we shall see,
one can readily formulate Kantorovich’s problem using this sort of software by
adding additional equations. We shall now show how you can use the package
lp_solve to reproduce Kantorovich’s solution to his problem.
The program requires that you input an expression to be maximised or min-
imised followed by a sequence of equations or inequalities. In Algorithm 2 we
give Kantorovich’s problem in the format that lp_solve requires. In this ex-
ample we use the following variables:
variable meaning
A number of units of A produced
B number of units of B produced
m1 number of milling machines used
m2 number of turret lathes used
m3 number of automatic turret lathes used
xij number of units of j produced on machine i
Thus x1a means the output of As on milling machines.
The first line of input is the objective function to be maximised. We give
this as A, meaning maximise the output of A’s. The following lines give the
constraints to which the maximisation process is to be subjected.

A-B=0

This is another way of writing that A=B, or that equal quantities

11 For example lp_solve and GLPK.

22
 of A and B must be produced.

m1<=3

This means that the number of milling machines used must be less
than or equal to 3. The characters ’<’ ’=’ are used because ≤ is not
available on computer keyboards. Similar constraints are provided
for the other machines.

m1-0.1x1a-0.05x1b=0
1 1
This specifies m1 = 0.1x1a+0.05x1b = 10 x1a+ 20 x1b or in words,
1
that allocating a milling machine to produce an A uses 10 of a
milling machine hour, and that allocating a milling machine to
1
produce a unit of B uses 20 of a milling machine hour. We provide
similar production equations for the other machines.

x1a+x2a+x3a - A=0

This says that the total output of A is equal to the sum of the out-
puts of A from each of the machines. We provide a similar equa-
tion defining the output of B.

Note that all equations have to be provided with variables and constants on the
left and a constant on the right. One can readily re-arrange the equations in this
form. The last line specifies that the number of units of A produced should be
an integer. When the equations are input to lp_solve it produces the answer:

Value of objective function: 86

Actual values of the variables:
A 86
B 86
x1a 26
x1b 6
x2a 60
x2b 0
x3a 0
x3b 80
m1 2.9
m2 3
m3 1

which exactly reproduces Kantorovich’s own solution (Table 4) arrived at using

his algorithm.

3.4 Generalising Kantorovich’s approach

In his first example Kantorovich deals with a very simple problem, producing
two goods in equal proportions using a small set of machines. He was aware,

23
even in 1939 that the potential applications of mathematical planning were
much wider. We will look at two issues that he considered which are important
for the more general application of the method.

1. Producing outputs in a definite ratio rather than in strictly equal quanti-

ties.

2. Taking into account consumption of raw materials and other inputs.

Suppose that instead of wanting to produce one unit of A for every unit of
B, as might be the case if we were matching car engines to car bodies, we
want to produce 4 units of A for every unit of B, as would be the case if we
were matching wheels to car engines ( and ignoring spare wheels). Can Kan-
torovich’s method deal with this as well. Consider Figure 1 again. In that the
plan ray is shown at an angle of 45◦ a slope of 1 to 1. If we drew the plan ray
at a slope of 4 to 1, the intersection with the production frontier would provide
the solution. Since this geometric approach only works for two products, let us
consider the algebraic implications.
You should now be convinced that it is possible to solve Kantorovich’s
original problem12 by algebraic means. In Algorithm 2 we specified that A −
B = 0 or in other words A = B , if one wanted 4 units of A for every B we
would have to specify A = 4B or, expressing it in the standard form used in
linear optimisation, A − 4B = 0 . Suppose A stands for engines, B stands for
wheels. If we now say wheels come in packs of 4, then we can repose the
problem in terms of producing equal numbers of packs of wheels and engines.
Introduce a new variable β = 4B to stand for packs of wheels, and rewrite the
equations in terms of β and we can return to an equation specifying the output
mix in the form A − β = 0, which we know to be soluble.
How do we deal with consumption of raw materials or intermediate prod-
ucts?
In our previous example we had variables like x1b which stood for the out-
put of product B on machine 1. This was always a positive quantity. Suppose
that there is a third good to be considered - electricity, and that each machine
consumes electricity at different rate depending on what it is turning out. Call
electricity C and introduce new variables x1ac, x1bc etc referring to how
much electricity is consumed by machine 1 producing outputs A and B. Then
add equations specifying how much electricity is consumed by each machine
doing each task, and the model will specify the total amount of electricity con-
sumed.
We now know how to :

1. Use Kantorovich’s approach to specify that outputs must be produced in

a definite ratios.

2. Use it to take into account consumption of raw materials and other in-
puts.
12 Actually this was his “problem A”

24
If we can do these two tasks, we can in principle perform in-natura calcula-
tions for an entire planned economy. Given a final output bundle of consumer
and investment goods to maximise and given our current resources, a system
of linear equations and inequalities can be solved to yield the structure of the
plan. From simple beginnings, optimising the output of plywood on differ-
ent machines, Kantorovich had come up with a mathematical approach which
could be extended to the problem of optimising the operation of the economy
as a whole.

3.5 A second example

Let us consider a more complicated example, where we have to draw up a plan
for a simple economy. We imagine an economy that produces three outputs
: energy, food, and machines. The production uses labour, wind and river
power, and two types of land: fertile valley land, and poorer highlands. If we
build dams to tap hydro power, some fertile land is flooded. Wind power on
the other hand, can be produced on hilly land without compromising its use for
agriculture. We want to draw up a plan that will make the most rational use of
our scarce resources of people, rivers and land.
In order to plan rationally, we must know what the composition of the final
output is to be - Kantorovich’s ray. For simplicity we will assume that final
consumption is to be made up of food and energy, and that we want to consume
these in the ratio 3 units of food per unit of energy. We also need to provide
equations relating to the productivities of our various technologies and the total
resources available to us.
Valleys are more fertile. When we grow food in a valleys, each valley
requires 10,000 workers and 1000 machines and 20,000 units of energy to pro-
duce 50,000 units of food. If we grow food on high land, then each area of high
land produces only 20,000 units of food using 10,000 workers, 800 machines
and 10,000 units of energy.
Electricity can be produced in two ways. A dam produces 60000 units
of energy, using one valley and 100 workers and 80 machines. A windmill
produces 500 units of electricity, using 4 workers and 6 machines, but the land
on which it is sited can still be used for farming.
We will assume that machine production uses 20 units of electricity and 10
workers per machine produced.
Finally we are constrained by the total workforce, which we shall assume
to be 104,000 people.
Tables 5 and 6 show how to express the constraints on the economy and the
plan in equational form. If we feed these into lp_solve we obtain the plan
shown in Table 7. The equation solver shows that the plan targets can best be
met by building no dams, generating all electricity using 541 windmills, and
devoting the river valleys to agriculture.
It also shows how labour should be best allocated between activities: 40000
people should be employed in agriculture in the valleys, 109 people should

25
 Table 5: Variables in the example economy
e total energy output
ec household energy consumption
f food
v valleys
w windmills
m machines
d dams
u undammed valleys
h highland
fh food produced on high land
fv food produced in valleys

Table 6: Resource constraints and productivities in our example economy

final output mix f = 3ec
number of valleys v=4
dams use valleys v−u = d
valley food output fv = 50000u
valley farm labour lv = 10000u
valley energy use ev = 20000u
valley farm machines mv = 1000u
highland food prod fh = 20000h
highland farm labour lh = 10000h
highland energy use eh = 10000h
highland farm machines mh = 800h
energy production e = 500w + 60000d
energy workers le = 100d + 4w
machines in energy prod me = 80d + 6w
workers making machines lm = 10m
energy used to make machines em = 20m
energy consumption em + ev + eh + ec ≤ e
machine use me + mh + mv ≤ m
total food prod f = fh + fv
workforce lm + le + lv + lh ≤ 104000

26
 Table 7: Economic plan for the example economy using lp_solve

d (dams) 0
e 270500
f 200218
h 0.0108889
m 6172.71
u 4
v 4
w (windmills) 541
ec 66739.3
eh 108.889
em 123454
ev 80000
fh 217.778
fv 200000
le 2164
lh 108.889
lm 61727.1
lv 40000
me 2164
mh 8.71111
mv 4000

work as farmers in the highlands, 2164 people should work on energy produc-
tion, and 61727 people should work building machines.
The results that we have obtained were by no means obvious at the out-
set. It was not initially clear that it would be better to use all the river valleys
for agriculture rather than building dams on some of them. In fact, whether
dams or windmills are preferred turns out to depend on the whole system, not
just on their individual rates of producing electricity. We can illustrate this by
considering what happens if we cut the labour supply in half to 52000 people?
If we feed this constraint in to the system of equations we find the optimal
use of resources has changed. The plan now involves 1 dam and 159 windmills.
Cut the working population slightly further, down to 50000 people and the
optimal plan involves flooding two valleys with dams and building only 23
windmills. Why?
As the population is reduced, there are no longer enough people available to
both farm the valleys and produce agricultural machinery. Under these circum-
stances the higher fertility of lowland valleys is of no importance, it is better to
use one or more of them to generate electricity. By applying Kantorovich’s ap-
proach it becomes possible for a socialist plan to do two things that von Mises
had believed impossible:

1. It allows the plan to take into account natural resource constraints - in this
case the shortage of land in river valleys which can be put to alternative
uses.

2. It allows rational choices to be made between different technologies - in

27
 this case between windmills and hydro power and between lowland and
highland agriculture.

Contrary to what von Mises claimed, the whole calculation can be done in
physical units without any recourse to money or to prices.

4 Valuation
The core of Mises’s argument relates to the use of prices to arrive at a rational
use of intermediate or capital goods. Mises argues that, in practice, only money
prices will do for this, but concedes that, in principle, other systems of valu-
ation, such as labour values would also be applicable. Kantorovich too, was
very concerned with the problem of relative valuation[26], and developed what
he called objectively determined valuations (ODV). These valuations differed
from prices, since a price involves an exchange of commodities for money be-
tween two owners. In the USSR all factories and all products were owned by
the state. When products moved from one factory to another, there was no
sale or purchase involved. The ODVs were purely notional numbers, used in
economic calculations, not selling prices.
He considered a situation where planners have to deal with several different
types of factories (A..E) each able to produce products 1 and 2, and where the
intended ratio of output of product 1 and 2 are fixed in the plan. Each class of
factory A..E has different relative productivities for the two products.
He next looked at the apparent profitability of producing products 1 and 2
under different relative valuations. Under some schemes of relative price, all
factories would find product 1 to be unprofitable relative to product 2, under
other the reverse would occur. Intermediate price schemes would allow both
products to be produced, with some classes of factories specializing on 1 and
others on 2. He gives the example of children’s clothing as something which,
under the administratively determined valuations then used in the USSR, were
unprofitable to produce, and unless factories were specifically instructed to
ignore local profitability, too few children’s clothes would be made.
He asks if there exists a relative valuation structure which would allow
factories to concentrate on the most valuable output, and at the same time meet
the specified plan targets and arrives at certain conclusions:

1. That among the very large number of possible plans there is always
an optimal one which maximises output of plan goals with current re-
sources.

2. That in the optimal plan there exists a set objectively determined valua-
tions (ODV) of goods which will ensure that each factory

(a) produces the output which will contribute most to maximising the
plan goals
(b) each factory also finds that the output which contributes to max-
imising plan targets is also the output which is most profitable

28
 3. With arbitrary valuations which differ from ODV, these conditions can
not be met, and profit maximising factories will not specialise in a way
that meets plan goals optimally.

It is important to understand that his ODVs are valuations that apply only for a
plan which optimally meets a specific plan target. Kantorovich’s procedure for
arriving at an optimal plan involved successive adjustments to the ODVs and
factory specialisation until both the appropriate mix of goods is reached, and at
the same time each factory is producing its most profitable good. He actually
gave several different mathematical procedures for arriving at such a plan and
system of ODVs.
Although Kantorovich asserts that labour is ultimately the only source of
value, his ODV’s are short term valuations and differ from the classical labour
theory of value, which gave valuations in terms of the long term labour re-
production costs of goods - including the reproduction costs of capital goods.
Kantorovich, in contrast, is concerned with valuations which should apply with
the current stock of means of production and labour resources. For example, he
considers the situation of giving a valuation to electric power relative to labour.
Instead of valuing it in terms of the labour required to produce electricity, he
first assumes that the total electrical power available is fixed - ie, power-stations
operating at full capacity, and then works out how many person hours of labour
is saved by using an additional kilowatt hour of electricity. The assumption is
made that in order to arrive at this objective valuation of electricity in terms of
labour

1. The plan targets must be met

2. The plan must be optimal

Kantorovich’s insistence on considering short term, very material constraints -

so many megawatts of power, such and such a number of cutting machines, etc,
gives his work an intensely practical and pragmatic character, quite different
from that of most theoretical economists.
Why is Kantorovich so concerned with valuations and profitability?
There seem to be two reasons. We should first note that by profit maximis-
ing Kantorovich actually meant maximising the value of output. This must be
understood in the context of Soviet practice where mines and factories were
given incentives to over-fulfill plan targets. If the output was a single good
- say coal, the target could be specified in tons. But if the factory produced
several goods, then the target had to be set in terms of x rubles worth of a mix
of goods. With the ’wrong’ price structure, plants would attempt to maximise
the production of the goods which were of the highest value, ignoring those of
lower value, with the result that the aggregate supply of all goods was often
not in the proportions that the planners intended. This practice of setting plan
targets in money terms reflected the limited ability of GOS PLAN to specify
detailed targets in kind as described by [43].

29
 The second reason relates to his particular algorithm for solving linear pro-
gramming problems which used an iterative adjustments to initial ODVs until
an optimal plan is achieved.
These two aspects seem intimately linked in his presentation, but the pre-
suppositions about the incentives to factories are not brought to the fore.
With computer algorithms, the process of solving a linear program becomes
a ’black box’. The user need not concern themselves with details such as the
method of calculation - whether it uses Kantorovich’s approach Danzig’s or
Karmarkar’s, except insofar as this affects the size of problem that can be han-
dled, as we discuss in section 5. With computer packages, ODVs would no
longer be needed for computing a plan, but would they still be needed for spec-
ifying targets to factories?
This depends on the information processing capacity of the planning sys-
tem. If it were capable of specifying fully disaggregated plans, then it could in
principle just place orders with factories for specific quantities of each good.
In these circumstances, the factories could not cheat by producing more of
high value items and less of low value ones. Indeed, the very information that
would be required to compute Kantorovich’s ODVs, would have been suffi-
cient for GOS PLAN to specify disaggregated orders in kind for the products
that would have had valuations attached.
There remains another level at which valuations would have been useful -
when product designs were being drawn up at a local level. If a refrigerator
designer was deciding on what components to use in a planned new model, she
would need some way of telling which components would, from a social stand-
point, have been the most economical, which implies a system of valuations.
However it is not clear that the full apparatus of ODVs would be either neces-
sary or appropriate here. ODV’s correspond to a system of marginal cost, rather
than average cost pricing. They reflect current marginal costs with the imme-
diately current constraints on production. The use of such marginal costing
was criticized by other Soviet economists[22, 37]. It is not clear, in retrospect,
that they would have been more appropriate than a system of average cost val-
uation if one was projecting ahead a year or so. Indeed, given the stochastic
properties of prices in a real capitalist economy[19], it is doubtful that, with
the exception of certain constrained products like oil, the difference between
average and marginal costs is significant in the west.

5 Complexity
Linear programming, originally pioneered by Kantorovich, provides an answer
in principle to von Mises claim that rational economic calculation is impossible
without money. But this is an answer only in principle. Linear programming
would only be a practical solution to the problem if it were possible, in prac-
tice, to solve the equations required in a socialist plan. This in turn requires
the existence of a practical algorithm for solving them, and sufficient compu-
tational resources to implement the algorithm. Kantorovich, in an appendix

30
to [25], gave a practical algorithm, to be executed by paper and pencil math-
ematics. The algorithm was sufficiently tractable for these techniques to be
used to solve practical problems of a modest scale. When tackling larger prob-
lems he advised the use of approximative techniques like aggregating similar
production processes and treating them as a single composite process. Whilst
Kantorovich’s algorithm uses his ODV’s, which he earlier called resolving mul-
tipliers, subsequent algorithms for linear programming do not, so the ODVs
should not be considered as fundamental to the field.
Since the pioneering work on linear programming in the 30s, computing
has been transformed from something done by human ’computors’ to some-
thing done by electronic ones. The speed at which calculations can be done
has increased many billion-fold. It is now possible to use software packages to
solve huge systems of linear equations. But are computers powerful enough to
be used to plan an entire economy?
In a large economy like the former USSR there were probably several mil-
lion distinct types of industrial products, ranging from the various sorts of
screws, washers and types of electronic components to large final products like
ships and airliners. Although there was great enthusiasm for Kantorovich’s
methods in the USSR during the 60s, the scale of the economy was to great for
his techniques to be used for detailed planning with the then available computer
technology. Instead they were used either in optimising particular production
plants, or, in drawing up aggregated sectoral plans for the economy as a whole.
How has the situation changed today, given that the power of computers has
continued to grow at an exponential rate since the fall of the USSR?

5.1 Complexity classes

To answer this one needs to be able to quantify the complexity of a planning
task and compare it with available computing resources. Measuring complexity
is a branch of algorithmics. Algorithms are classified into complexity classes.
For instance, computing the average of a list of n numbers is said to be of com-
plexity class n, because the number of simple arithmetic operations required
will be proportional to n. This complexity class is termed linear, as the algo-
rithms execution time on a computer grows linearly with the number of items.
A bit more complex than linear algorithms are the log-linear ones. It turns
out that one can sort a list of n numbers into ascending order using n log(n)
basic arithmetic operations. Problems which are either linear or log-linear are
reckoned to be very easy to solve on computers.
Next in difficulty come the polynomial problems where the number of basic
arithmetic steps grows as some polynomial function of the size of the input
data. If an algorithm had a running time that was proportional to n2 or to
n3 for some size of input data n then it would be of polynomial complexity. In
algorithmics, polynomial problems are regarded as being tractable, since, with
computers able to do billions of operations per second, such problems can be
solved for quite large values of n. For example multiplication is a task that

31
grows polynomially with the number of digits in the numbers. If you want to
multiply 17 by 32 you have to carry out the basic steps 2 × 7 = 14, 2 × 10 = 20,
30 × 7 = 210, 30 × 10 = 300 and then add up the partial products. The number
of multiplication steps will grow as n2 , where n is the number of digits in the
numbers.
Beyond the polynomial problems comes the class of NP or non-deterministic
polynomial problems. These are problems which, were you to take them to Or-
acle at Delphi, and were the priestess to give you an answer, you could check
whether her answer was correct in polynomial time. Suppose you had a 100
digit number x and asked the priestess what its prime factors were, and she
replied with one 47 digit number and one 53 digit number. You could take
this on trust, or bearing in mind the many tales of those mislead by the Divine
Oracle, you might decide to check if her answer was correct. If she were right,
then multiplying the two numbers she gave you should yield x. This multipli-
cation would take you of the order of 47 × 53 = 2491 basic operations, which is
roughly 14 n2 in terms of the length of the original number you gave the priest-
ess. This shows that we can check the validity of purported prime factors in
polynomial time.
Sadly, the Oracle at Delphi has long fallen from use, and we, lacking that
divine guidance once available to the Ancients, must find prime factors by
mundane means. A mundane and deterministic procedure is to test all prime
√
y ∈ 2.. x to see if yx is a whole number. The first such y is a prime factor. The
drawback is the vast number of tests that must be performed. For 100 digit
numbers we would have to test all y in the range 2..1050 to be sure of finding a
n
prime factor if one existed. The number of tests to be performed grows as 10 2 ,
in other words the number of tests grows exponentially with n. This problem,
and others in the class of exponential problems, is assumed to be computa-
tionally intractable, since the number of possibilities to be checked grows so
rapidly that it rapidly exhausts the power of even the swiftest computer. Indeed
so hard is the task that certain cryptographic protocols[49] rely on large prime
factors being practically impossible to discover.

5.2 Complexity class of economic planning

After that short introduction to the idea of complexity classes, let us apply
these ideas to economic planning. To what complexity class does linear pro-
gramming belong?

For a long time it was not known whether or not linear pro-
grams belonged to a non-polynomial class called “hard” (such
as the one the traveling salesman problem belongs to) or to an
“easy” polynomial class (like the one that the shortest path prob-
lem belongs to). In 1970, Victor Klee[29] and George Minty cre-
ated an example that showed that the classical simplex algorithm
would require an exponential number of steps to solve a worst-
case linear program. In 1978, the Russian mathematician L. G.

32
 Khachian[28] developed a polynomial-time algorithm for solving
linear programs. It is an interior method using ellipsoids inscribed
in the feasible region. He proved that the computing time is guar-
anteed to be less that a polynomial expression in the dimensions
of the problem and the number of digits of input data. Although
polynomial, the bound he established turned out to be too high for
his algorithm to be used to solve practical problems.
Karmarkar’s algorithm [27] was an important improvement on
the theoretical result of Khachian that a showed how linear pro-
gram can be solved in polynomial time. Moreover his algorithm
turned out to be one which could be used to solve practical linear
programs. ( Dantzig [11])

Modern linear programming packages tend to combine Dantzig’s simplex method

with the more recent interior point methods. This allows the most modern
implementations to solve programming problems involving up to one billion
variables[21, 20]. For such huge problems large parallel supercomputers with
over a thousand processor chips are used. But even with much more modest
4 CPU computers, linear programming problems in the million variable class
were being solved in half an hour using interior point methods13.
These advances in linear programming algorithms and in computer technol-
ogy mean that linear programming could now be applied to detailed planning
at the whole economy level, rather than just at an aggregate level.

6 Deriving the plan ray

Kantorovich assumed that the plan had a given target to optimise in the form of
a particular mix of goods: the plan ray. This reflected the social reality for those
engaged in managing Soviet industry, in that they were given a mix of prod-
ucts to produce by GOS PLAN. The planning authorities themselves however,
needed to specify what this ultimate output mix would be. In the early phases
of Soviet planning, when Kantorovich wrote his original paper, the goals set
by the planners were primarily directed at achieving rapid industrialization and
building up a defence base against the threat of invasion. The planning pro-
cess was successful in achieving these goals. But in an already industrialised
country, in times of peace, the meeting of current social needs becomes the first
priority and so plan vector has to be pointed in that direction. A criticism com-
monly leveled at the Soviet-type economies—and not only by their Western
detractors—is that they were unresponsive to consumer demand. It is therefore
important to our general argument to demonstrate that a planned economy can
be responsive to the changing pattern of consumer preferences—that the short-
ages, queues and surpluses of unwanted goods of which we hear so much are
not an inherent feature of socialist planning. The economists Dickinson and
13 See[3] chap 4. The Harmony Algorithm for constructing plans, given in [9], is an instance of

the class of algorithm discussed by Bienstock.

33
Lange, writing just prior to Kantorovich, outlined a practical mechanism by
which this could be done [31, 16].
They proposed that the state wholesale sector should operate on a break-
even basis with flexible prices. Wholesale managers would set market clearing
prices for the products on sale as consumer goods. These wholesale prices
would then act as a guide to the plan authorities, telling them whether to in-
crease or decrease production of particular lines of product. If prices were high,
then that line of product would have its output increased, otherwise its planned
output would be reduced.
The basic idea is clear, the same principle that adjusts production of con-
sumer goods in a capitalist economy was to be employed. But this then raises
the problem of how one determines that a price is high or low. High or low
relative to what?
What would be the basis of valuation used?
After incorrectly rejecting the possibility of planning in kind, Mises had
considered the possibility that the socialist planners might be able to make use
of an ‘objectively recognizable unity of value’, i.e., some measurable property
of goods, in performing their economic calculations. The only candidate Mises
could see for such a unit is labour content, as in the theories of value of Ricardo
and Marx. The latter had proposed that workers be paid in labour tokens and
that goods be priced similarly[33]. Mises ended up rejecting labour as a value
unit; he had two relevant arguments, each purporting to show that labour con-
tent cannot provide an adequate measure of the cost of production. These argu-
ments concern the neglect of natural resource costs implicit in the use of labour
values, and the inhomogeneity of labour. Mises’s critique of labour values is
very brief and sketchy. Two pages or so of substantive argument appear in [58]
and are reproduced in [60]. This doubtless reflects the fact that although Marx
and Engels had laid great stress on planning as an allocation of labour time,
this conception had been more or less abandoned by English speaking socialist
economists by the late 30s. Neither Lange nor Dickinson relied on the classical
theory of value in their arguments. Writing in 1930, Appel[1] had laid great
stress on the relevance of the labour theory of value for socialist economics, but
his ideas were largely ignored. More recent writers have again laid emphasis
on Marx’s theory of value as a guide to socialist planning[17, 46, 45, 9].
The basic principle in these schemes can be stated quite simply. All con-
sumer goods are marked with their labour values, i.e. the total amount of so-
cial labour which is required to produce them, both directly and indirectly. But
aside from this, the actual prices (in labour tokens) of consumer goods will be
set, as far as possible, at market-clearing levels. Suppose a particular item re-
quires 10 hours of labour to produce. It will then be marked with a labour value
of 10 hours, but if an excess demand for the item emerges when it is priced at
10 labour tokens, the price will be raised so as to (approximately) eliminate the
excess demand. Suppose this price happens to be 12 labour tokens. This prod-
uct then has a ratio of market-clearing price to labour-value of 12/10, or 1.20.
The planners record this ratio for each consumer good. We would expect the

34
ratio to vary from product to product, sometimes around 1.0, sometimes above
(if the product is in strong demand), and sometimes below (if the product is
relatively unpopular). The planners then follow this rule: Increase the target
output of goods with a ratio in excess of 1.0, and reduce the target for goods
with a ratio less than 1.0.
The point is that these ratios provide a measure of the effectiveness of so-
cial labour in meeting consumer’s needs (production of ’use-value’, in Marx’s
terminology) across the different industries. If a product has a ratio of market-
clearing price to labour-value above 1.0, this indicates that people are willing
to spend more labour tokens on the item (i.e. work more hours to acquire it)
than the labour time required to produce it. But this in turn indicates that the
labour devoted to producing this product is of above-average ’social effective-
ness’. Conversely, if the market-clearing price falls below the labour-value,
that tells us that consumers do not ’value’ the product at its full labour content:
labour devoted to this good is of below-average effectiveness. Parity, or a ratio
of 1.0, is an equilibrium condition: in this case consumers ’value’ the product,
in terms of their own labour time, at just what it costs society to produce it.
The feasibility of using labour time for expressing prices depends on being
able to calculate it. This might seem a daunting task, but it actually involves
solving a similar, though somewhat easier, set of linear equations to those re-
quired when one draws up a consistent plan. The task is thus computationally
tractable on the grounds explained earlier.
Mises objected that "the ... defect in calculation in terms of labour is the
ignoring of the different qualities of labour" (1935: 114). Mises notes Marx’s
claim that skilled labour counts as a multiple of, and hence may be reduced
to, ‘simple labour’, but argues that there is no way to effect this reduction
short of the comparison of the products of different labours in the process of
market exchange. Wage differentials might appear to offer a solution, but the
equalizing process in this case "is a result of market transactions and not its
antecedent." Mises assumes that the socialist society will operate an egalitarian
incomes policy, so that market-determined wage rates will not be available as a
guide to calculation. The conclusion is then that "calculation in terms of labour
would have to set up an arbitrary proportion for the substitution of complex
by simple labour, which excludes its employment for purposes of economic
administration" (1935: 115).
True, labour is not homogeneous, but there is no warrant for the claim that
the reduction factor for complex labour has to be arbitrary under socialism.
There are two possible approaches:

1. Skilled labour may be treated in the same way that Marx treats the means
of production in Capital, namely as a produced input which ’transfers’
embodied labour to its product over time. Given the labour time required
to produce skills and a depreciation horizon for those skills, one may
calculate an implied ‘rate of transfer’ of the labour time embodied in
the skills. If we call this rate, for skill i, ri , then labour of this type

35
 should be counted as a multiple (1+ri ) of simple labour, for the purpose
of ’costing’ its products. An iterative procedure is needed: first calculate
the transfer rates as if all inputs were simple labour, then use those first-
round transfer rates to re-evaluate the skilled labour inputs, on this basis
recompute the transfer rates, and so on, until convergence is reached.

2. Alternatively one may use the approach advocated by Kantorovich[26](page

64..66) where he shows that skilled labour of different grades can be as-
signed ODVs on the basis of their different productivities.

Which method is used would depend on the timescale of the calculation. If one
wants short term answers to the relative valuations of different labours, then
Kantorovich’s approach is relevant. For longer term considerations, within the
time scale that newly trained staff can be brought up to speed, then the first
alternative would be appropriate.

7 Conclusion
The Soviet mathematical school founded by Kantorovich and the Austrian
school exemplified by Mises and Hayek took radically different positions on
the feasibility of socialist economic calculation. To a large extent they ignored
one another. The Austrian school largely concentrated on criticising Western
trained socialist economists like Lange and the Soviet school appears to have
ignored Mises completely. Even when the key participants met, the issue was
not raised. Menshikov writes:

It is interesting that in the account of his trip to Sweden for

receiving the Nobel Prize, Kantorovich mentions an informal re-
ception with the participation of several American economists –
Nobel Prize laureates – including Hayek, Leontief, and Samuel-
son. But, apparently, neither at this reception, nor during other
meetings, this issue was never raised. In January 1976, when I
worked in USA as the Director of the United Nation Projections
and Perspective Studies Branch, I was asked to present L. V. Kan-
torovich as a new Nobel Prize laureate at the annual meeting of the
American Economic Association in Atlantic City. Of course, I put
the emphasis on the economic discovery of the laureate. In the dis-
cussion, none of the audience, which included T. Koopmans and L.
Klein, a future Nobel Prize laureate, mentioned the question of ac-
tual Kantorovich’s answer to a part of Hayek’s argumentation.[37]

With the political demise of the USSR, the Austrian school have tended to
assume that Mises arguments have been vindicated, but theoretical economic
arguments are not finally resolved by politics. Political fashions change. So-
cialism, from being politically unpopular in Europe the 1990s, has, since then,
been making substantial inroads on another continent. No, one has to bring

36
economic arguments head to head in their own terms. Kantorovich, an ab-
sent participant in the Western debate on socialist calculation, is worth paying
attention to.

References
[1] J. Appel, M. Baker, A.A. Deutschlands, and G.I. Kommunisten. Fun-
damental Principles of Communist Production and Distribution. Move-
ment for Workers’ Councils, von der Kollektivarbeit der Gruppe Interna-
tionaler Kommunisten - GIK [Allgemeine Arbeiter Union Deutschlands
- AAUD], 1990 (1930).

[2] Aristotle. The Politics. Hutchinson, 1988.

[3] D. Bienstock. Potential function methods for approximately solving lin-

ear programming problems: theory and practice. Kluwer Academic Pub-
lishers, 2002.

[4] M.L. Bierbrier. The Tomb-Builders of the Pharaohs. American Univ in

Cairo Press, 1989.

[5] Paul Cockshott and Allin Cottrell. Information and economics : a critique
of hayek. Research in Political Economy, 18(1):177–202, 1997.

[6] W. P. Cockshott and A. F. Cottrell. Labour time versus alternative value

bases: a research note. Cambridge Journal of Economics, 21:545–549,
1997.

[7] W. P. Cockshott and A. F. Cottrell. A note on the organic composition of

capital and profit rates. Cambridge Journal of Economics, 27:749–754,
2003.

[8] W.P. Cockshott and A. Cottrell. The Scientific Status of the Labour The-
ory of Value. IWGVT conference at the Eastern Economic Association
meeting, in April, 1997.

[9] Allin Cottrell and Paul Cockshott. Towards a New Socialism, volume
Nottingham. Bertrand Russell Press, 1992.

[10] Allin Cottrell and Paul Cockshott. Calculation complexity and planning :
the socialist calculation debate once again. Review of Political Economy,
5(1):73–112, January 1993.

[11] G Dantzig. Linear Programming. Operations Research, 50(1):42–47,

2002.

[12] G. B. Dantzig. Linear Programming and Extensions. Princeton Univer-

sity Press, Princeton, 1963.

[13] A.R. David. The Pyramid Builders of Ancient Egypt:: a Modern Investi-
gation of Pharaoh’s Workforce. Routledge, 1996.

37
[14] R. Dawkins. The God Delusion. Houghton Mifflin Co., 2006.

[15] R. Dawkins et al. Unweaving the Rainbow: Science, Delusion and the
Appetite for Wonder. Houghton Mifflin Company, 1998.

[16] HD Dickinson. Price Formation in a Socialist Community. The Economic

Journal, 43(170):237–250, 1933.

[17] H. Dieterich. La democracia participativa: el socialismo del siglo XXI,

País. Vasco, Baigorri, 2002.

[18] Michel-Stephane Dupertuis and Ajit Sinha. A sraffian critique of the

classical notion of center of gravitation. Technical report.

[19] Emmanuel Farjoun and Moshe Machover. Laws of Chaos, a Probabilistic

Approach to Political Economy. Verso, London, 1983.

[20] Jacek Gondzio and Andreas Grothey. Massively parallel implementa-

tion of interior point methods for very large scale optimization. In
R. Wyrzykowski, J. Dongarra, N. Meyer, and J. Wasniewski, editors, Par-
allel Processing and Applied Mathematics, Lecture Notes in Computer
Science, 3911, Berlin, September 2006. Springer-Verlag.

[21] Jacek Gondzio and Andreas Grothey. Solving nonlinear financial plan-
ning problems with 109 decision variables on massively parallel archi-
tectures. In C. A. Brebia, editor, Computational Finance and its Appli-
cations II, WIT Transactions on Modelling and Simulation, volume 43.
WIT Press, 2006.

[22] Gregory Grossman. Review: Against bourgeois economic pseudo-

theories of socialism. The American Economic Review, 53(1):211–213,
1963.

[23] KE Iverson. A programming language. Wiley, New York, 1966.

[24] Kenneth E. Iverson. Notation as a tool of thought. In ACM Turing award

lectures, page 1979. ACM, New York, NY, USA, 2007.

[25] LV Kantorovich. Mathematical Methods of Organizing and Planning Pro-

duction. Management Science, 6(4):366–422, 1960.

[26] L.V. Kantorovich. The Best Use of Economic Resources. Harvard Uni-
versity Press, 1965.

[27] N. Karmarkar. Anew polynomial-time algorithm for linear programming.

Combinatorica, 4(4):373–395, 1984.

[28] L. Khachian. A polynomial algorithm in linear programming. Soviet

Mathematics Doklady, 20:191–194, 1979.

[29] V. Klee and G. Minty. How good is the simplex algorithm. In O. Shisha,
editor, Inequalities-III, pages 159–175. Academic Press, 1972.

38
[30] H.D. Kurz and N. Salvadori. VON NEUMANN’S GROWTH MODEL
AND THE’CLASSICAL’TRADITION. Understanding" classical" Eco-
nomics: Studies in Long-Period Theory, 1998.

[31] Oscar Lange. On the Economic Theory of Socialism. University of Min-

nesota Press, 1938.

[32] B. Lumpkin. Geometry Activities from Many Cultures. Walch Publishing,

1997.

[33] K. Marx. Marginal Notes to the Programme of the German Workers’

Party [Critique of the Gotha Programme]. Marx and Engels Selected
Works, 3, 1970.

[34] Karl Marx. The difference between the democritean and epicurean phi-
losophy of nature. In Marx-Engels Collected Works Volume 1. Progress
Publishers, 1841.

[35] Karl Marx. Capital, volume 1. Progress Publishers, Moscow, 1954. Orig-
inal English edition published in 1887.

[36] S. Meikle. Aristotle’s Economic Thought. Oxford University Press, 1997.

[37] S.M. Menshikov. Topicality of Kantorovich’s Economic Model. Journal

of Mathematical Sciences, 133(4):1391–1397, 2006.

[38] G. Michaelson, W. P. Cockshott, and A. F. Cottrell. Testing marx: some

new results from uk data. Capital and Class, pages 103–129, 1995.

[39] J. Neumann. A Model of General Economic Equilibrium. Review of

Economic Studies, 13(33):1–9, 1945.

[40] O. Neurath. Economic plan and calculation in kind. Otto Neurath: Eco-
nomic Writings 1904-1945, 2004.

[41] Otto Neurath. The Conceptual Structure of Economic Theory and its
Foundations. In Thomas Uebel and Robert Cohen, editors, Economic
Writings. Kluwer, (1917) 2004.

[42] Otto Neurath. Economics in Kind, Calculation in Kind and their Rela-
tion to War Economics. In Thomas Uebel and Robert Cohen, editors,
Economic Writings. Kluwer, (1919) 2004.

[43] Alex Nove. The Economics of Feasible Socialism. George Allen and
Unwin, London, 1983.

[44] J. O’Neill. Markets, Socialism, and Information: A Reformulation

of a Marxian Objection to the Market. Social Philosophy and Policy,
6(2):200–210, 1989.

[45] A. Peters. Computersozialismus:Gespräche mit Konrad Zuse. Vaduz,

2000.

39
[46] Arno Peters. Das Äquivalenzprinzip als Grundlage der Global
Ökonomie. Akademische Verlagsanstalt, 1996.

[47] S. Pinker. The Language Instinct: How the Mind Creates Language.
HarperCollins, 2000.

[48] R. Remak. Kann die Volkswirtschaftslehre eine exakte Wissenschaft wer-

den. Jahrbücher für Nationalökonomie und Statistik, 131:703–735, 1929.

[49] RL Rivest, A. Shamir, and L. Adleman. A Method for Obtaining Digital

Signatures and Public-Key Cryptosystems. Communications, 1978.

[50] L.D. Ryabev and Y.N. Smirnov. The Atomic Project, Science, and the
Atomic Industry. Atomic Energy, 99(2):519–527, 2005.

[51] A. M. Shaikh. The empirical strength of the labour theory of value.

In R. Bellofiore, editor, Marxian Economics: A Reappraisal, volume 2,
pages 225–251. Macmillan, 1998.

[52] Piero Sraffa. Production of commodities by means of commodities. Cam-

bridge University Press, Cambridge, 1960.

[53] A. M. Turing. Computing Machinery and Intelligence. Mind, (49):433–

460, 1950.

[54] A. M. Turing. Lecture on the Automatic Computing Engine, 1947 . In

B. J. Copeland, editor, The Essential Turing. OUP, 2004.

[55] Alan Turing. On Computable Numbers, With an Application to the

Entscheidungsproblem. Proceedings of the London Mathematical Soci-
ety, 42:230–65, 1937.

[56] Thomas Uebel. Incommensurability, Ecology,and Planning: Neurath in

the Socialist Calculation Debate, 1919–1928. History of Political Econ-
omy, pages 309–342, 2005.

[57] CJ Van Rijsbergen. The Geometry of Information Retrieval. Cambridge

University Press, 2004.

[58] L. von Mises. Economic calculation in the socialist commonwealth. In

F A Hayek, editor, Collectivist Economic Planning. Routledge and Kegan
Paul, London, 1935.

[59] L. Von Mises. Socialism: An Economic and Sociological Analysis.

Johnathan Cape, 1951.

[60] Ludvig von Mises. Human Action. Hodge and Company, London, 1949.

[61] J. von Neumann. Mathematical Foundations of Quantum Mechanics,

Engl. transl. of the 1931 German edition by RT Beyer, 1955.

[62] D. Widdows. Geometry and Meaning. Number 172 in CSLI Lecture

Notes. University of Chicago Press, 2004.

40
[63] David Zachariah. Testing the labor theory of value in sweden.
http://reality.gn.apc.org/econ/DZ_article1.pdf, 2004.

[64] David Zachariah. Labour value and equalisation of profit rates. Indian
Development Review, 4(1):1–21, 2006.

41