Calculation Problem Again

Justin Schwartz jschwart at
Sun Oct 30 14:41:51 MST 1994

On Sun, 30 Oct 1994 tgs at wrote:

> Justin,
> Mandel and McNally have both written critiques of the "we can't have enough
> information" line.  The capitalists themselves now plan how they stock their
> distributors vis centralized planning with decentralized inputs.

Of course there is planning, quite a lot of it, under capitalism.
Corporations are themselves planned systems. It does not follow from the
viability of direct allocation in corporations or negotiated allocation
among firms in both cases in the context of a market system that direct,
nonmarket allocation can replace market allocation at an acceptable level
of efficiency.

The way Mandel tries to solve the information problem--McNally really just
gestures at it and dismisses it as bourgeois ideology--is twofold. First,
he says we can reduce the amount of information by simplying the product
mix: talk about shirts instead of t-shirts vs sports shirts, etc. (You can
see that this would work wonders with consumer satisfaction.) Second he he
says "most" demand is inelastic--turns out he means 65%-70% of it, doesn't
say where he gets this figure--and we can plan for that percentage by
projecting on last year's figures. Notice that this still leaves you with
an economy that is 35%-40% marketized, so Mandel in fact doesn't really
offer a wholly planned alternative, or pretend to. But even for the
planned part Mandel is stuck without any way to deal with changes in
demand. Given the "simplification" solution to the information problem to
start with at the static level (this "solution" Hayek called part of the
problem,!) and the rigidity at the dynamic level, you have a recipe for
massive inefficiency and a guaranteed emergence of a huge black market.

  As for
> their being too many planners in a decentralized system, or even too many
> inputs, this is basically Hobbes-warmed over.  that is to say: only if you
> posit human beings as being incapable of meeting together and establishing
> some consensus via majority decision-making in popular assemblies, only if
> you think of us as inherently striving for power after power and a-social
> and a-moral and inherently competitive do you come up with this notion that
> too many cooks spoil the broth.  But man's political and both social nature,
> as Aristotle discusses in the Politics and Marx in the Grundrisse, is the
> basis for positing the possibility of any kind of socialism.

No no no no no no. In the first place the calculation problem does not in
any way turn on people being selfish. The problem with too many planners
to which I was alluding is just that if it's not possible for say 300
people in bureaucracy to know everything you need to know to plan
effectively, it is no more possible for 150 million people to know all
that stuff too. In fact the more you spread the info around the more it
will degrade by a process of "telephone." Altruism won't help that.

Democratic decisionmaking is not that easy either. It's not clear to me,
for example, why a majority vote for a plan which, when it is done,
reflects none or few of my first choices for anything, and on which I have
had a 1 in 150 millionth part in forming and choosing, counts as anything
more democratic than voting with my dollars in the market. And there are
all the usual problems with aggregating individual preferences into a
social choice.

Finally, any economic program that is premised on the idea that
self-interest will vanish is doomed. Hobbes had a point. We don't need to
warm him over. He's still hot. I note again that the calculation problem
does not depend on Hobbesean assumptions. It would arise in a community of

The best attempt at answering the calculation problem of which I know is
Albert and Hahnel's and it isn't good enough.


More information about the Marxism mailing list