Alchian and Demsetz: "Production, Information Costs, and Economic Organization"

Could be titled: "The Firm as a Highly Specialized Surrogate Market"

What characterizes firms?

Not authority to coordinate by fiat. ("This is delusion.")

Firms/entrepreneurs/managers do not own all of their inputs -- particularly human capital.

They must provide incentives such that this capital is used efficiently in production.

This is done through contracts, just as it is between individuals and firms in what we normally consider market transactions.

A&D argue that "the absence of the price mechanism/presence of director of inputs" is not what firms are about.

They are organizational forms characterized by the team use of inputs plus the centralized position of some party in the contractual arrangements of all other inputs.

Relationship to Coase

Can be interpreted as extension.

Change "cost of using the price mechanism" v. "cost of contracting through markets" to "cost of contracting through this particular class of institutions"

A&D push further, though, and explain why contracting through this class of institutions might be less expensive, and under what circumstances we would expect it to be so.

Answer: shirking, shared inputs (non-separabilities)

When production is more efficient with shared inputs than non-shared ones, it may be more efficient to establish sets of agreements that characterize firms than to govern these transactions using other institutional situations: for example, markets.

Examples of: "Team use of inputs"

Examples of "centralized position of some party in contractual arrangements" In some cases, shared use of inputs is more efficient than non-shared use. "Whole more than sum of the parts."

The Metering Problem: suppose individual output is extremely difficult or impossible to directly observe: only collective output can be.

Incentive problem: benefits from effort are shared by the whole group, but costs are borne individually.

This gives rise to shirking, unless individuals within group can costlessly detect and punish.

(This is why I never liked group projects.)

Individual contracts among individuals in this case can be extremely costly.

Possible solution: external competition to deter shirking

outsiders bidding to replace shirkers (lower share)

But how can they tell who is shirking? And won't they shirk just as much when they are employed? They face similar incentives.

Another solution: hire a monitoring specialist.

allow them to contract directly with each input, right to terminate, renegotiate contracts -- a right which those not party to these contracts don't have, let them sell their right in the market.

**need to be able to infer individual productivities from observables reasonably well (although this need not be verifiable) -- otherwise there will not be efficiency improvements.

Who monitors the monitor?

The market, if the monitor is the residual claimant -- reaps the marginal benefits from increased productivity.

Ownership arises from role as monitor! (We might think the direction of causation is the other way around...)

Efficiency gains benefit all of those in the organization.

We have now characterized the classical firm!


A&D argue that these will arise as a consequence of arranging an efficient contracting structure in cases where shared/team use of inputs increases productivity.

"Agency theory of the firm."