Coase's central proposition is that the difference of system operating cost (transaction cost) leads enterprises to replace the market. On the one hand, market transactions involve products or commodities; On the other hand, "enterprise transaction" involves factors of production. The growth of enterprises is regarded as the substitution of factor market for product market, which leads to the saving of transaction costs. This proposition is not easy to understand, because Coase does not define "enterprise", and as we will see, there is no clear distinction between product market and factor market.
Assume that the production input is private. Therefore, every input owner is faced with three choices: (1) producing and selling goods by himself; (2) simply selling his production inputs; (3) enter into a contractual arrangement to entrust the right to use the input to the agent in exchange for income. The emergence of enterprises is related to the third choice: entrepreneurs or agents, who hold a limited set of use rights stipulated in the contract to directly carry out production activities without directly referring to the market price of each activity, and then take the products so produced to the market for sale. There's a problem here. If private property rights do not exist, then the above options are not applicable. It is relatively easy to understand that the flow of a worker or an input is determined by the agent rather than the market price. But why should private property owners voluntarily give up their rights and be at the mercy of a visible hand?
According to Coase, this choice is to reduce transaction costs. Some people think that this statement is synonymous repetition. But this is not the case, because people also put forward other reasons to explain the emergence of enterprises, including division of labor, risk and production cooperation. Coase considered these factors and denied them all. In his view, transaction costs constitute the main part of the consideration. His argument is easy to refute, because other seemingly reasonable factors may lead people to speculate that the total transaction cost may rise when enterprises appear.
Even if transaction cost becomes the only relevant factor, or even if other factors are strongly interpreted as subordinate to the scope of transaction cost, if someone can identify different types of transaction costs and how they change in different situations, we still can't say that Coase's argument is tautological. Extreme generalization will lead to tautological argumentation, and too little universal applicability will lead to concrete argumentation. Between the two, we can find testable meaning, and how to determine the specific cost is a matter of choice, depending on the problem at hand.
Emphasizing transaction costs does not deny the potential benefits of specialized division of labor or effective cooperation in production efforts. For example, consider the classic "pin factory", where each owner of various inputs focuses on only one part of his work. If the total transaction cost is zero, customers who buy pin will have to pay each of many contributors separately. Comparative advantage guides everyone to specialize in their own skills. If it seems advantageous to hire a coordinator, the buyer of the pin can simply pay the extra fee to this person. In this case, a large number of product prices command the production of a needle.
In such a world, it will be redundant to talk about product market and factor market. The two are inseparable: the buyer pays the contribution of both the product and the owner of the input. To separate the product from the factor market, the agent needs to pay the owner of the input and get the reward from the consumer for handing over the product. The standard research assumes that the quantity of products is given, and Coase's view is that this quantity can only be determined after the transaction cost is explicitly included in the analysis.
We can explain this problem more fundamentally. Any production input is private property. If it is within the clearly defined boundaries, its owner has the right to: (1) exclude others to ensure that he can independently decide its use; (2) the right to exclusive use proceeds; (3) the right to transfer or exchange property (including labor) with anyone he thinks fit. This kind of exchange right means contractual right, and property rights can be traded through various contractual arrangements. Of course, the exchange of rights, for the owner, is to obtain higher income, and the choice of contract has transaction costs.
Here, our main concern is not the transaction itself, but the contractual arrangement, from which the right to use the input is authorized to the other party, and the goods produced by using the input are sold to consumers. Not completely transferred, because the owner of the input reserved some other rights, and the contract became a structured document. A set of defined right to use will be transferred in exchange for income. In the form of contract, the input owner will be bound to obey orders, rather than deciding his own policies by constantly referring to the market prices of various actions he may take.
The transfer of the right to use is a matter of degree, and the definition of the right to grant is a matter of contracting, which is often supplemented by implied understanding, custom and common law. Needless to say, a clerk will not be asked to do the work left to the janitor. The remuneration paid is usually based on measurable attributes (such as how many hours you work in a day), which is not similar to the measurable attributes on which the final commodity sales pricing is based. Therefore, the observable market price cannot directly guide the behavior of the input owner, just as every action can be measured and priced. Therefore, the transfer of the right to use usually means granting the right to decide what to do.
As an axiom, economic analysis asserts that under private property rights, when an input owner enters the above contract form (joins the enterprise), he expects the benefits relative to other options, because he could have chosen not to join the enterprise. What must be explained is why there has been an increase. It is usually more productive to let another person make a decision. The answer is no, we notice that the granting of decision-making power is caused by the difference of pricing and measurement attributes. Generally speaking, management decision-making cannot be superior to consumer decision-making realized through price mechanism. When price information guides every activity, mistakes will be reduced.
Then, will the specialization, cooperation and planned economy realized by many owners bring higher income to everyone, so that everyone chooses to join the enterprise? The answer is no again. As mentioned above, if every activity can be measured and priced, then the benefits of specialization and collaboration can be realized without the "factor market"-the right to decide and use someone's factors does not need to be granted to agents or entrepreneurs. Because in the product market, every contribution of the factor owner will be paid.
Do enterprises appear because people are lazy, cheating or speculating, as some recent papers have said? Maybe. But the problem is that this kind of behavior is common, but the degree and nature change, depending on the selected contract method or the measurement and pricing method of property in the transaction. The lazy behavior of factory workers needs the services of supervisors, which is the result of his granting himself the right to use labor. If every small contribution of a worker is rewarded, he will not be lazy, or at least his laziness will be very different.
Coase's answer is bold: "The main reason why entrepreneurship is profitable seems to be the cost of using the price mechanism. The most obvious cost of' organizing' production through the price mechanism is to find out what the relevant price is. " (The emphasis is added by the author. He put forward some reasons for the high cost of finding and negotiating, although his explanation seems incomplete. I want to put forward four general reasons, at least two of which come from Coase.
Perhaps, the most obvious reason for the higher price of no enterprise is to conduct more large-scale transactions, and each transaction should be priced separately. If customers pay for every contribution or component of a product instead of a finished product, the cost will often be too high to trade. As an alternative, all cooperative input owners can enter into a contract with another person, and everyone agrees to pay the price for his service, and then all these prices can be added to the price of the final product. Instead of these contracts, the central agency can sign a contract with each input owner, pay him the right to use and sell the final product at another price. On this way of reducing contracts, Coase wrote: "One factor of production (or its owner) does not have to establish a series of contracts with multiple factors in the enterprise, just as, of course, if this cooperation is the direct result of the operation of the pricing mechanism, because this series of contracts has been replaced by one contract."
The second factor that Coase obviously didn't consider was to understand the information cost of products. When it is not easy to determine the parts of a product separately, it may be much more expensive for producers and consumers to reach an agreement on the price of each part than on the price of the whole product. The price of a spring in the camera is disproportionately higher than the cost of reaching an agreement on the price of the camera. Although the consumer has the final decision on the value evaluation of the whole product, it is impossible to expect him to identify the value of every part of the product-he doesn't even know what some of them are or whether they exist. Everything we buy has to be calculated clearly, which is simply too expensive. We will see that in piece-by-piece contracts, the parts of goods are often directly measured and priced, but these negotiations are conducted between professional agents and input owners, that is, as far as a part itself is concerned, it is not easy to determine its value. It is cheaper to reach an agreement on price between professionals and input owners than between owners and consumers or between professionals and consumers. Manufacturers of parts know these parts better than their consumers.
The third cost of finding a price is to measure the cost. In every transaction, whether the transaction is between the agent and the customer, between the agent and the input owner, or between the input owner and the customer, several characteristics or attributes must be measured. If the activities of the owners of inputs change frequently, if the scope of these activities changes greatly, or some upcoming activities cannot be specified in advance, it is often more economical to give up the direct measurement of these activities and replace them with another measurement method. In this way, agents can hire people by the hour or rent houses by the square feet without having to evaluate every contribution of input factors. In fact, all kinds of input activities as the source of product value are not priced at all, because the evaluation fee is too low. According to the property completely different from the actual contribution of the owner of the input or the property of the final product sold to consumers, the measurement shall be made and the remuneration shall be paid to the owner of the input. Agents involve two different ways of measuring: to bear the profits and losses by guiding and supervising the work of the input owners, and to provide consumers with the overall goods of a specific quality.
When Dancausse wrote the following words, he seemed to deliberately use "alternative" measurement methods to avoid measuring the cost of various auspicious activities:
"For those who provide services or goods, it may make no difference which course of action to take, but it is not the case for the buyers of services or goods. But the buyer doesn't know which route he will take for the supplier. Therefore, the service to be provided is only a general expression, and the specific details will be dealt with later. All the terms in the contract are the boundaries of what the commodity or service provider should do. The details of what the supplier is expected to do are not written in the contract, but decided by the buyer in the future. When the flow of resources (within the scope stipulated in the contract) becomes dependent on the buyers who adopt this method, we can obtain what I call the "enterprise" relationship ... which is obviously more important in the case of service labor than in the case of goods. Taking commodities as an example, it is of little significance to specify the main projects in advance and then decide the details later. "
Finally, in the process of reaching a price agreement, there is a cost for the decomposition of contributions. In cooperation with the owner, in some cases, everyone's contribution may not be easy to describe, and everyone will ask for more than his due share. In fact, competition among input owners will reduce excessive demand, but the problem has not been solved. Therefore, agents can exercise rights similar to dominance by hiring cooperative participants, and once again provide prices for everyone's stay by measuring the substitution index rather than the contribution itself.
To illustrate the benefits of collaboration and the difficulty of decomposing contributions, Archin and Demsetz gave examples of loading and fishing. My favorite example is the towing of inland river shipping in China before the establishment of the * * * production party regime, in which a large group of workers pulled wooden boats in the river with neat steps along the river bank. The uniqueness of this example is that the collaborators actually agreed to hire a foreman to whip them. The point here is that even if every tracker is completely "honest", it is still expensive to measure everyone's contribution to the ship's movement. However, it is difficult to get the unanimous consent of all people by choosing different measurement methods, so the arbitration of agents is essential.
For this example, Archin and Demsetz believe that the supervisor (that is, the enterprise) participates because of laziness. My argument goes further, asserting that this behavior itself is produced because it measures substitute indicators rather than efforts, so the concept of laziness is an indirect way to express the cost of finding relative contribution prices. Therefore, in Coase's so-called enterprise, laziness occurs in a different way from deception in the market.
Let me summarize the argument of this lengthy part. In principle, like the services of partners, all the contributions of input owners can be priced and sold to customers by directly measuring various attributes related to each contribution. In this case, the product market and the factor market overlap. However, because of the number of transactions, because consumers lack detailed information about the use of each component or component of a commodity, because of the variety and difficulty in changing, and because of the need to decompose contributions, price determination is expensive.
An effective way to reduce the cost of price discovery is to replace the direct and separate pricing of activities with some settings. This alternative method can be as simple as adopting a piece-rate wage contract, or as complicated as establishing a * * * producer system. Coase's main concern is that it needs an alternative way for private input owners to grant the right to use within the scope that everyone's behavior or contribution cannot be directly priced. This means that through the use of alternative indicators, "entrepreneurs" have emerged.
Here, specialization and collaboration are related to scope, they usually involve multiple input owners, and tend to mix the cost of finding prices. Without the cost of measuring and evaluating performance, there would be no enterprise and the value of social output would be maximized. But these expenses do exist, and when they are partially replaced by factor market contracts, the transactions in the product market will be reduced. When it is found that the cost of price is reduced, the agency cost of foreman, director or manager-also the transaction cost-will rise. When the saving of one cost is equal to the marginal increase of another cost, substitution stops.
The typical product market in textbooks is composed of direct sales of goods and services, and its price is determined by the relationship between supply and demand. It does not involve the contractual arrangement of paying remuneration to producers (suppliers) according to alternative indicators rather than product measurement. Factor market appears in another chapter, typically under the marginal productivity theory. The income obtained by each input owner is equal to the value of its marginal product, not to mention the authorization of the agent. In fact, it doesn't seem to matter how and through whom the input owners get paid.
It is incorrect to say that "enterprise" has replaced "market". Rather, one form of contract replaced another. Coase's main concern is a form of contract, that is, the owner of the input gives up a set of clearly defined rights to use his input in exchange for income. So he was commanded by an invisible hand, not by the invisible hand of the price mechanism. Extraordinary insight will show that when this form of contract increases, the product market will decrease.