Arnold Kling  

More Thoughts on Market Socialism

PRINT
NCAA Fesses Up--in Prime Time... Is Transparency the Answer?...

My first thoughts were here. Below are some further thoughts.

1. Does market socialism suffer from information problems? It would seem that government calling out prices to profit-maximizing firms might be no worse than market prices from an information standpoint. According to Lawrence White, even Mises reverted to saying that the problem was incentives. Mises suggested that even if you could present socialist managers with a list of shadow prices, without the ability to share in profits the managers would tend to ignore prices and do what is easiest.

2. Perhaps you could solve the incentive problem. For example, the Masonomic view (and that of Adam Smith) is that people care about status. Suppose that a society can award status points to managers who maximize "profits" based on shadow prices given out by the government.

3. Clearly, one thing that government cannot know is subjective valuation. With government-set prices, how does anyone figure out that some people like to play golf and others like to folk dance?

4. A market socialist would counter with, "how does the market know social benefit?" With market-set prices, how does anyone figure out that automobiles cause pollution or that the social benefit of $10,000 of medical services for Joe is higher than having that amount of services go to Fred?

5. A challenge for market socialism is to figure out what is a final good and what is an intermediate product. For example, if movies and video games are final goods, then government has to fix the right relative price between the two. If the final good is "entertainment," then government can just fix the price of entertainment and let the market determine the relative prices of movies and video games. Similarly, there is a challenge in distinguishing fundamental inputs from intermediate outputs. Is human capital a fundamental input, or is it an intermediate output that results in part from schooling decisions and other choices?

6. If under market socialism the government only sets prices for final outputs and fundamental inputs, then new products and new technologies can emerge just as in a market economy (assuming the incentive problem is solved by issuing status points). However, if it is too hard to come up with workable definitions and measures of final outputs and fundamental inputs, then new products and new technologies must be valued by central planners, which will create severe information problems.

7. Going back to a question I raised in an earlier post, why do large firms rely primarily on command and control, rather than using shadow prices? We can see the use of something like shadow prices when firms use formulas to determine compensation, including sales commissions, worker bonuses, and executive stock option awards. One flaw with such formulas is that they can be gamed. Just as Soviet managers tended to game the quotas that came from the planning board, any clever salesman can game a commission structure. As I have argued in the past, management has to constantly tinker with incentive structures in order to stay one step ahead of employees learning to game the system.

8. It seems to me that within a firm, the valuation of what Garett Jones calls organizational capital is highly arbitrary and subjective. To put this another way, senior management has great difficulty articulating exactly what each employee contributes to the organization. In fact, I think that our current high unemployment rate reflects the fact that in many organizations senior management decided arbitrarily and subjectively to reduce their implicit valuation of what their employees were doing. It would be extremely difficult and time-consuming for management to come up with a shadow price on the output of its accountants, marketing departments, research and development units, and so on. It is quicker and easier to use command-and-control.

I think that is what the socialist calculation debate boils down to. Setting shadow prices from the top down is an extremely difficult and time-consuming process. If you have the power, it is much easier to exercise it using command-and-control techniques than to try to articulate your goals by using shadow prices. Your shadow prices will inevitably be incompletely specified, which leaves you subject to being gamed by managers lower down the chain of command.

Markets set prices much more quickly and responsively than any central planner can do. The problem, from a socialist point of view, is that market prices are a seemingly arbitrary function of individual subjective preferences, rather than an expression of what is "known to be correct for society as a whole."

Market socialism is an intellectual attempt to avoid the trade-off between having an efficient, real-time pricing system on the one hand or reverting to crude command and control on the other. But in fact there is no middle ground between the two.

In my view, if you centralize power, you inevitably pay a price in terms of having an inferior information flow. In that regard, I side with the Austrians in the socialist calculation debate. I think that the recent trend toward greater centralization of power here in the United States is one for which we will pay a large price. Of all the phenomena that represent potential threats to our quality of life, the hubris of our ruling class is the one that concerns me the most.


Comments and Sharing





COMMENTS (10 to date)
Robert Johnson writes:

"...even Mises reverted to saying that the problem was incentives. Mises suggested that even if you could present socialist managers with a list of shadow prices, without the ability to share in profits the managers would tend to ignore prices and do what is easiest."

This seems completely right to me, based on my observation of managers in my own company. It's hard work to find and implement the best solutions. With a weak incentive (the possibility of recognition for success) to do so, few people do, and the ones who do are those who are either interested in the problem for personal reasons (e.g. engineers like new tech, so they research it), or who have been given a strong incentive, like avoiding public failure on a high-profile project.

Allan Walstad writes:

"Markets set prices much more quickly and responsively than any central planner can do."

Agreed, that's the main point. But why is it the main point? Because what counts is progress. Given fixed factor inputs, fixed technology, and a fixed bundle of final goods, let us grant that the socialist planners could eventually settle on the most effective allocation of the inputs to maximize output. Meanwhile, in the dynamic capitalist economy, entrepreneurs are finding ways to apply new technology and introducing new goods. Some souces of raw materials become relatively depleted, others are discovered, tastes change, populations grow or shrink and change demographically, people move, etc, etc. There's no way for the socialist planners to keep up, so things get frozen.

The only reason socialist calculation could have ever been taken seriously is because mainstream econ got hung up on equilibrium models, arguably out of the desire to maintain a technical scientific status for the field, you know, with all sorts of fancy math. The Austrians didn't get hung up on equilibrium models.

Grant writes:

4. A market socialist would counter with, "how does the market know social benefit?" With market-set prices, how does anyone figure out that automobiles cause pollution or that the social benefit of $10,000 of medical services for Joe is higher than having that amount of services go to Fred?

How does the government know the social benefit?

If the costs (including transaction and coordination costs) of providing a social good are less than the benefits, a market will provide that good. Is there a similar mechanism in states, and if so which ones? Most people would say yes, but while its obvious that states provide public goods, it isn't clear to me that their public goods outweigh their costs (which include provided public bads, such as wars of aggression).

david writes:

@7: Firms use command and control because industries where shadow prices make sense are also industries that simply break up into smaller constituent firms that sell directly to each other on the open market, rather than a subsection of said market.

Ultimately someone has to be making the bets; you can't keep adding layers of market abstraction and expect the quality of information to stay the same.

TK writes:

Some of the deficiencies pointed out above can be handled by allowing managers to keep their profits, and using taxes on incomes to improve social benefit. The taxes can also be spent in a decentralized way, by substituting direct goverment spending with programs like school vouchers, or direct income transfers. This would still acheive the socialist goal of reducing inequality along with the advantages of distributed decision making on what resources are spent on. Essentially, what Friedman & others pointed out long ago.
Also, use Pigou taxes for goals which are good for society but individuals dont have an incentive to spend on. Of course, there is still the crucial question of how to decide what these goals should be, if any. But given some set of goals, one could still acheive them without centralized resource allocation.
Eventually, one would hit the peak of the Laffer curve and further taxes would mean less resources for acheiving these goals. But this is a constraint faced in any resource allocation scheme.

Joe Teicher writes:

The firms that I have worked for made extensive use of shadow prices. In addition to compensating based on an individual or group's profits, they would determine profits using internal prices. So, for instance, you would pay a desk fee for each unit of office space you consumed, with that fee covering the cost of the actual space, plus some amount for use of communal food and beverage, accounting, electricity and networking and IT support. If you wanted to have the internal pool of software developers do something for you, they would get billed out at an hourly rate and that would be a charge against your P&L.

I know this wouldn't work in many industries or for many jobs, but in the prop trading industry its standard.

fundamentalist writes:

The real problem of socialist calculation is waste. The only purpose of production is consumption. If production produces things that people don't want, then those resources are wasted. For socialism to waste less than free markets, the state planners would have to be able to discern the wants of consumers better than the consumers themselves. Not likely. Hence greater waste.

Economic progress happens when we produce more than we waste. Even in a laissez-faire market there is a great deal of waste.

Socialists don't like what consumers want. They think they know better than consumers what they should be consuming. But we already have many historical example of attempts at such. What the state produces for the consumers' "good" no one wants and the waste piles up. At the same time, a black market develops that produces what consumers want.

How many centuries do we have to rehash this debate? The historical evidence is very clear about what happens when the state tries to force consumers to buy what it thinks is best for them. We have dozens of examples spanning a century! Why the continuing debate?

James writes:

Arnold,

Re: point 2, I'm sure you believe that the government can create adequate incentives to get most citizens to adhere to the minimum wage, to obey obey quantity controls on prescription drugs, etc. Why then do you think that governments are so incapable of incenting their citizens to follow the orders of an economic planning committee?

Governments can credibly threaten to kill workers who don't play ball. They can lock citizens in iron cages for years on end. Govenments can make it a legal requirement to hold a government issued id card to be eligible to work and threaten to take that id card away from noncompliant persons. Governments can issue ration cards for the purchase of food and withold those cards from anyone who won't get with the program. Are such threats not incentive enough?

Real socialist governments actually used all of these to incent their victims. Do you think that the socialist experiments of the 20th century failed because the incentives created by such threats were still too weak to get people to follow the plan?

Tracy W writes:

It would seem that government calling out prices to profit-maximizing firms might be no worse than market prices from an information standpoint.

How rapidly could a government call these prices out? What would happen if a set of prices needed to change suddenly, eg an earthquake wipes out a significant chunk of the world's aluminium supply? Remember it's not just the price of aluminium that needs to change, but the price of all of the potential substitutes for aluminium, and the substitutes for those substitutes and so forth.

Dyer1829 writes:

I have to agree with the issue of final goods. How does the government determine if one movie is better than another? Who would decide that? Not to mention that some people like movie "A" better than movie "B", but other people like movie "B" better than movie "A". The government has no definitive way to pick which movie or game or anything that is based on entertainment is better than any other.

I know this is not brought up, but I would like to tack on one extra point. If the government sets prices, what happens to competition? Why would anyone want to make a better product when they can sell the one they already have for the exact same price? No one will want to, development will cease and competition will be eliminated.

Comments for this entry have been closed
Return to top