[lbo-talk] how a non-market economy would work - WAS Re: socialist response to hayek

SA s11131978 at gmail.com
Wed Apr 1 10:32:54 PDT 2009


Philip Pilkington wrote:


>> Allin F. Cottrell and W. Paul Cockshott, "Information and Economics:
>> A Critique of Hayek."
>> http://www.reality.gn.apc.org/econ/hayek.htm
>> _
>>
>
> That really is a fine piece of writing. Was I the only one though that was a
> little disconcerted by the authors' argument. I don't mean this from any
> logical or rational point of view - indeed their article was extremely tight
> and one wouldn't be hard pushed to see something of the sort eventually come
> into existence, this especially in light of the disastrous results of
> today's potent combination of misinformation and human stupidity.
>
> What really irked me though was their discussions of the "single mind".

[....]


> Now the theory being put forward
> here seems to be something like a technocratic society run by a giant
> computer which itself ran along the lines of perfect
> exchange/identity/equivalence. Thus after dismissing Hayek's "director"
> figure they sneak him in the back door... in the form of some sort of giant
> computer!
>
> This raises an awful lot of questions and not merely cultural ones, like
> what such a society would look like and how desirable living in it would be.
> It also, in my mind at least, raises questions such as: what exactly happens
> when you have a computer that follows two basic axioms: exchange and growth?
>

Cockshott and Cottrell published a book in 1994, _Toward A New Socialism_, where they laid out how this system works in detail. I just went back and re-read/re-skimmed it. It really is a remarkable work. The authors are mathematicians and/or computer scientists, and they've obviously studied the history and theory of economic planning quite thoroughly. Their book fearlessly faces up to all the difficult issues involved in running a non-market economy, with a minimum of handwaving, and comes up with what really seem - to me, at least - like the best possible solutions.

They've put the entire book online, by the way, here: http://www.ecn.wfu.edu/~cottrell/socialism_book/

As a public service, and to illustrate what a non-market economy might look like (or might have to look like), below I've written a brief summary of how the Cockshott/Cottrell system works. It's a full-length, heavily detailed book and I've done this hastily, so I may have misunderstood or misrepresented some points. (Corrections welcome.)

---

1. All values are expressed as labor-hours. One hour of labor is rewarded with one labor-token. However not all labor has the same "labor content": One hour of highly educated or trained labor embodies "more labor" than one hour of less trained labor (due to the labor costs of training). Thus, each worker is assigned an index value according to his degree of education, for planning purposes. A worker with average education has an index value of 1.0. A worker with above-average education - say, a graduate degree - might have an index value of 1.3. Below-average labor might have a value of 0.8. Differently educated workers do not receive different levels of pay. If a labor shortage arises in a particular occupation, the authors are reluctant to accept higher pay rates as incentives (though that's always an option). They seem to prefer either simply cutting back on planned production requiring those types of labor, offering non-monetary incentives (more control over work, e.g.), or "directed labor," which is what it sounds like. However, people are paid differently according to effort, which is graded by firms. (There are, e.g., three grades of labor, A, B, and C, depending on effort.)

2. The Plan begins with the desired "net outputs" - i.e., the quantities of final outputs to be consumed by individuals. Then the Plan works backward, calculating the quantities of intermediate outputs (products used in the production of other products) that will be needed to produce these. It seems that the desired final outputs are determined by a combination of two processes. One is "strategic planning," which is rather vaguely defined. But the strategic plan might, among other things, include a stipulation that, say, 7% of national resources will go to production of consumer electronic goods. Apparently, this level of planning is considered too detailed to submit to popular vote (unlike broader planning targets such as percentage spent on health, education, investment, etc.) It is determined democratically by "planning juries" selected at random from the population. The other determinant of final output is consumer demand, described below in point #4.

3. Planning calculations are facilitated by two databases. First, there is a Universal Product Code database, which assigns a unique numerical code to each unique product, down to the level of size, style, type, etc. (So, Cheez-O brand cheese puffs has one code; Cheez-Man brand puffs has a different code; and Cheez-Man brand bbq flavored, bite-sized cheese puffs has a third code.) There are expected to be tens of millions of unique products, including intermediate goods (i.e., a 2.5mm steel-alloy rotary lathe aperture has its own code). Then, for each UPC-coded product, the database documents the technology - how much labor and precisely how much of each intermediate good - identified by UPC - is needed to produce each UPC-coded product.

For each product, a labor-value (comparable to a capitalist price) must be calculated. As with capitalist prices, labor-values must be constantly broadcast and updated in order to maintain overall economic efficiency. It's done this way: Each firm has a computer with a spreadsheet program, which communicates with the central planning computer. Each period, the firm manager enters into the spreadsheet how much inputs it used up - how much labor and how much of each intermediate product (identified by UPC code) - as well as how much output it produced. Of course, to translate this information into the numerical labor-value for a product, the labor-values of each of the intermediate products must be known. Thus, all this information is communicated by all firms, for all products, to the central planning computer. The computer calculates the labor values for all products by integrating all this information into a gigantic input-output matrix. Then the central computer broadcasts (via "teletext," we would now say Internet) the constantly updated labor-values of each product back to the firms. That way, each firm can see which intermediate products are getting more "expensive" and which are getting cheaper, and adjust production technique accordingly. (I think that's the point, anyway.)

4. A Marketing Authority (MA) is in charge of all retail/wholesale - i.e., getting products to consumers. Consumer goods are allocated according to a simulated market mechanism. I found this the cleverest part of the scheme. The MA orders products wholesale from firms. Each product (identified by UPC) has an up-to-date labor value, calculated as described above. Then the MA distributes the products to the stores, where consumers can choose to buy them, using their labor-tokens. The MA adjusts the retail prices to reflect supply and demand - if a product is flying off the shelf, its price is gradually raised. This means that the retail price of a good can diverge from its underlying labor-value. For example, if the labor-value of a shoe is 10 labor-hours, but it turns out to be popular among consumers and its price needs to be increased to 12 labor-hours (labor tokens) to balance supply and demand, then it has a price/value ratio of 1.2. For any product with a ratio greater than 1, orders must be given to the producer to increase production; for goods with ratios below 1, orders must be given to reduce production.

5. The marketing system is important because it also allows the performance of enterprises to be evaluated: A firm that overstated its input requirements would produce a product with a relatively high labor-value, which would result in a relatively low price/value ratio. If a different firm had a higher ratio for the same product, the laggard firm with the low ratio could be easily identified. So it's clear which consumer-goods firms should be "rewarded" with more orders. However, this only works for consumer goods, where there is a simulated market. There is no market for intermediate and capital goods. So to estimate the effectiveness of capital-goods producers, a statistical sampling technique is used: For a given capital good, a list is drawn up of all the consumer goods produced using the capital good. Then the price/value ratios of those consumer goods are collected. Since the list of consumer goods is basically a "random sample" of all consumer goods, statistically one should expect only a certain level of deviation of the price/value ratios from the average ratio. If the deviation is above-normal, in one direction or the other, that's a clue that the capital-good producer may be operating inefficiently.

6. The authors only briefly discuss innovation, within the context of "strategic planning." They seem to see innovation as a function of aggregate R&D spending and urge the establishment of "institutes" devoted to developing new technologies. They write that "for R&D to be effective there must be a transmission belt that spans the stages of pure research, applied research, product development, and mass production." This will apparently be carried out by the research institutes.



More information about the lbo-talk mailing list