Chapter 1

Make or Buy?

The make-or-buy decision is a classic management concern. Every firm uses thousands of inputs, and for each there is a potential to either manufacture the input itself or acquire it on the market. In its broadest interpretation, this decision includes choices like hiring a consultant or employing internal labor to perform a given task. If a firm decides to make an input, it will transact internally with a division or another part of the firm. If it decides to buy, it will contract with another organization. In either case, it is important to understand the principles behind the structure chosen and behind the transaction. The make-or-buy decision is sometimes treated as an accounting or financial decision. While it is obviously important to perform accounting analyses and to choose the low-cost method, it is more important to understand the managerial basis of the decision.

One advantage of such understanding is that it will economize on decision-making time. There are virtually thousands of products that a firm will use, and each of them could potentially be produced internally. Managers need some method of deciding which products are good candidates for internal production and are worth a detailed internal cost analysis. Of course, some of the decisions are obvious -- a messenger firm will not go through a complex analysis in order to decide whether or not to build its own cars. But other decisions are less obvious -- if the messenger firm gets large enough, should it make its own uniforms or buy them? (Answer: Buy them.) The principles in this chapter will enable a manager to quickly and easily eliminate a whole host of products from consideration and analysis as potential products to make internally. (As seen below, the analysis of the messenger firm and its uniforms was an immediate consequence of the principles set forth in the remainder of this chapter.) This itself will save considerable managerial and accounting time.

The make-or-buy decision is often analyzed in terms of either capital market issues or technology. As indicated later, analysis of make-or-buy choices cannot be done correctly if only technological or financial criteria are considered, although of course both are important to the decision. Rather, the correct analysis is a management analysis. The decision should depend on particular managerial elements, and the purpose of this chapter is to elucidate these elements.

USE OF THE MARKET

Subject to certain availability constraints, the firm wants to acquire inputs as cheaply as possible. When a competitive open market exists, this usually offers the most powerful method of controlling costs. If a product is made internally, then the firm must spend substantial managerial resources monitoring costs and efficiencies. In the market, on the other hand, simple shopping or seeking bids can easily and cheaply control costs. The best way to control costs is through the market. Provision of inputs by sellers facing competition provides powerful obvious incentives for low-cost, reliable production. Therefore, the first presumption should always be for purchasing inputs on the market. This conclusion should not be startling. After all, most firms buy most of their inputs, from office furniture to paper clips to automobiles to steel. It is only in special cases that the firm vertically integrates and makes its own inputs.

The analysis of the make-or-buy decision should therefore depend on examining factors which interfere with market provision of the inputs. Ultimately, these factors have to do with the cost and availability constraints mentioned above. However, the issue is not the existence of these constraints, but rather the reasons why the constraints may come into play. We must ask why the market will not provide the product cheaply, or why something may happen to make the input unavailable.

AVOID EXPLOITATION

The answer turns out to rest on the possibility that suppliers will be in a position to exploit the firm. This occurs when buying an input will subject the firm to aholdupproblem which will potentially exist when the firm is subject toopportunistic behavior.When this happens, suppliers will be in a position to increase prices of inputs, or to demand additional payments before making needed inputs available in a timely manner.

Opportunistic behavior may occur when one firm can take advantage of its position with respect to another firm, as either a customer or a supplier. If a manager puts his firm in a position where someone can exploit it, the manager has laid the firm open to the risk of being the victim of opportunistic behavior. To avoid this, it is important to understand conditions under which this kind of behavior can occur.

The important point to keep in mind here is that there is a difference between the return needed on an investment before it is undertaken and the smaller return needed to maintain and operate an investment once it has been finished and the sunk costs are sunk. A manager will undertake an investment only if she can predict a return high enough to cover the firm's cost of capital. After the investment is undertaken it will often pay to continue to operate it at a rate of return which would not have been high enough to justify undertaking the investment in the first place. This difference -- the difference between the amount which would be needed to justify an investment and the amount which justifies operating it after it is undertaken -- is potentially available to be exploited from the firm. (Economists have given this quantity the particularly unpleasant namequasirents.) A good manager will be aware of the possibility of having quasirents exploited from his firm and will not place the firm in such a position for this to occur.

For an example which shows what is involved, consider a firm planning to get into the municipal garbage business. There are two major parts of the business -- inputs needed to pick up the refuse and a landfill for disposal. First, the firm has to consider the purchase of trucks and bid for a contract in a city. The current price for landfill capacity is $5 per ton and the firm plans to dispose of 1 million tons per year if it receives the contract. Trucks will cost $5 million (scrap value of $2.5 million) and the firm must invest $1 million in learning the local market and establishing contracts. Both the trucks and the goodwill of the initial investment will last fifteen years. Therefore, the amortized cost of this investment at 10 percent is about $800,000 per year. Drivers, maintenance, and other variable costs and the profit needed to make the endeavor worthwhile are $1.2 million per year. Therefore, for the 1 million tons involved, the firm must be able to charge at least $7 million ($5 million for landfilling, $1.2 million for variable costs, $800,000 return on fixed costs), or $7 per ton. (See Table 1.) If it cannot charge at least this amount, it will not bid on the contract.

But woe betide the manager who bids the job at $7 per ton and buys a fleet of trucks on the basis of these calculations. He has laid his firm open to severe exploitation by the owner of the landfill. The landfill owner can raise his price to $5.45 per ton and the truck operator will be better off paying it than shutting down. If he shuts down, he can sell the trucks for $2.5 million but he will still owe the rest of the cost of the trucks, another $2.5 million, plus the $1 million invested in establishing the business. If he sells the trucks and shuts down, he will still owe $460,000 per year to pay off the debt incurred in buying the trucks and setting up the business. If he continues to operate, he will lose $450,000 per year. In other words, the truck company loses $10,000 more by shutting down than by operating. Note, however, that the company would never have gone into the trash business if it had expected to pay $5.45. The $.45 per ton represents the quasirent which can be exploited from the company by the landfill operator behaving opportunistically.

CONTRACTUAL SOLUTION?

A manager potentially facing this situation could try to draw up a contract in advance specifying the landfill price at $5.00 per ton. A sensible businessman would want such a contract, but it may not be enough. (In the next chapter, we discuss ways of structuring a transaction which will give the maximum protection whenvertical integration-- here a decision to buy the landfill -- is not feasible.) A general principle is that all complex contracts are incomplete. It is impossible to specify a contract which will cover all eventualities.

First, of course, in times of inflation the landfill owner would not agree to a contract with a price fixed in money terms, so the contracting process becomes more complex. This issue, however, creates no particular difficulties. Prices may be tied to some general index, such as the Consumer Price Index, or to a narrower index of input prices. (In landfilling, it might be an index related to the minimum wage, the price of gasoline, and truck prices. This index will be more appropriate for this transaction than would be a general price index such as the Consumer or Producer Price Index, but since it will have to be designed and calculated specifically for this contract it will also be more expensive.)

The second, more substantial, difficulty is writing a contract with enough specificity to avoid any opportunistic behavior. While it may not appear so, landfilling is nonetheless a complex task. There are many ways in which the landfill can "hold up" the truck company for additional revenue. For example, it can make the trucks wait longer for their turn to dump their trash, thus increasing costs borne by the hauling company. It can also allow the company to use only relatively inaccessible parts of the landfill. It would be very difficult to specify the contract in enough detail to eliminate any possibility for exploitation. As environmental concerns regarding landfilling increase, and as liability for toxic waste disposal becomes more of a problem, holdup possibilities also increase.

This is not a legal problem. The difficulty is not that the lawyers are not clever enough to draw up a contract. If the businessman can tell them what types of opportunistic behavior might be involved, the lawyers can probably eliminate these behaviors by contract. The attorneys might even be able to suggest additional types of opportunistic behavior to avoid contractually -- if they have seen other contracts or been party to litigation regarding such contracts or holdup situations.

But there remains a business problem. It is unlikely that a businessman can be clever enough to think up all the ways in which a landfill could exploit its position, and the lawyers will not have been exposed to contracts or litigation involving all forms of opportunism. Some events may be of such low probability that it does not pay to spend the resources to anticipate their occurrence, to decide what should happen if they occur, or to draft the rules which are appropriate for achieving these goals. (Most likely, more contracts in California than in Iowa specify the consequences of earthquake damage.) In certain situations, contracts will not be adequate to eliminate all possibilities of opportunism because they are incomplete in some way, and it may not be possible or worthwhile to make them complete

Even if a contract could be written with sufficient specificity, enforcement would be difficult. In many of the cases where exploitation is possible, time is important, so that the delay associated with litigation over a contract breach might itself impose substantial costs on the firm. In the garbage example, landfills by definition fill up, so that by the time the truck company might win a suit, there may be no space left for dumping. Whatever damages can be collected may not be adequate to compensate for the lost quasirents.

VERTICAL INTEGRATION

What is the solution? As the story is told, the best solution is for the truck company to buy the landfill before bidding the hauling contract. (Actually, an option on the landfill contingent on getting the municipal hauling business would probably be the chosen method.) If the same company owns the landfill and the trucks, there is no possibility for exploitation of the quasirents. This is a case where the firm should clearly consider seriously "making," not "buying." Ownership, what economists callvertical integration,is the answer. The landfill and the trucking company are worth more under joint ownership than they are worth if owned separately.

What features of this example make ownership compelling? The answer is the close relationship between the capital needed for trash hauling and the capital needed for landfilling Once the city contract is awarded, the hauling business is dependent upon the landfill. The landfill is aspecific assetfor the hauling business. The two sets of assets are closely linked, and therefore a possibility for opportunistic behavior exists if they are separately owned.

This dependence, however, may be symmetric. It is quite possible that the hauling company will be able to hold up the landfill. This is particularly true if the hauling company is one of the major users, or the sole user, of the landfill. Just as the landfill is a specific asset for the truck company, the trucking business is a specific asset for the landfill. The hauling company can perhaps offer too low a payment to the landfill, threatening to use another landfill further away, and exploit some of the quasirent invested by the landfill owner.

The key point is not that vertical integration will save haggling costs; it is much more fundamental. If holdup possibilities exist, then efficient transactions may not occur. Socially and privately desirable investments may :not be undertaken, and some efficient deals will not be done.

The possibility of holdup will deter the landfill owner as well as the truck company from undertaking otherwise desirable investments. For example, if the truck company asks the landfill to undertake some additional investment, such as building new access roads, the landfill owner will hesitate because this will then give the truck company a holdup possibility over the landfill. After the road is built, there are quasirents associated with it, and these can be exploited by the truck company just as easily as the quasirents on the truck business can be exploited by the landfill. The trucks are a specific asset for the landfill, just as the landfill is a specific asset for the trucks.

If the landfill and the truck company are owned by the same party, only then will neither party be subject to holdup. Common ownership will eliminate the possibility of opportunistic behavior being used to exploit quasirents. Moreover, it does not really matter who buys whom. The landfill owner could just as easily own the truck company as the other way around. The only important point is that both inputs be under common ownership. When the two assets are under joint ownership, all efficient investments will be undertaken. When they are not, some efficient deals will not be clone. Some economists would even say that separate ownership can lead to market failure.

The two assets together are worth more if they are owned by one party than if they are separately owned, and this is the fundamental requirement for vertical integration. It also means that efficient ownership structures should be observed in the market. The hauling company can pay more for the landfill than anyone else because of this increased value of joint ownership.

It is the combination of exploitable quasirents (which is a common business phenomenon) and someone in a position to exploit them (a less common event, and one to be avoided) which should lead to such vertical integration. It is only when both features are present that vertical integration is useful. For instance, it would not pay the hauling company to buy a truck factory because there is no "asset specificity" between a truck manufacturer and a hauling firm. There are many competitive bidders for contracts to sell trucks, and if one truck company tried to hold up the hauling company, the latter could easily turn to another. (Since there are many clothing manufacturers in a position to make uniforms to any desired specification with no delay, there is no holdup possibility, which is why the messenger firm discussed above need not vertically integrate into uniform manufacture.)

Note that the form of business organization chosen will also effect the type of assets used. We may think of alternative production processes. One involves specific assets and is more productive; the other uses only general, nonspecialized assets but is less productive. (A more productive process involving general assets would dominate, and so need not be considered since if it is available it will always be chosen.) Then which process: is chosen will depend in part on the structure to be used. If there is to be vertical integration, then it becomes efficient to invest in more specific assets and use the more efficient process. Thus, there is an interrelationship between type of capital used and the organizational structure of the firm, and a manager should realize that both decisions should be made simultaneously.

NOT MONOPOLY

The possibility of being held up is not the same as being a victim of a monopolist. A monopoly position is not required for quasirent exploitation. Before the truck company bids on the hauling contract, there are many potential bidders. None has any monopoly power. Nonetheless, after a bid is successful, the bilateral monopoly position between the hauling company and the landfill exists. What was a competitive position before actual investment now becomes a noncompetitive position, precisely because the investment, creates exploitable quasirents. This switch -- from a competitive position before investment in undertaken to a bilateral monopoly after the investment -- is sometimes called thefundamental transformation.

Consider a situation in which there is only one landfill but many hauling companies. The landfill will charge a high, monopoly price to all hauling companies. Hauling companies will make their plans and undertake investments on the basis of this high price. But the monopolist landfill will not raise the price after the hauling company has undertaken its investment in order to exploit the quasirents. If the landfill did this, then future entrants into hauling would quickly learn of this behavior and refrain from entering, reducing demand for the services of the landfill. A monopoly position will lead to high prices, but it takes more than this to generate exploitable quasirents and thus the incentive for vertical integration.

It is inappropriate to use the antitrust laws :in situations of asset specificity, even though these situations are characterized by bilateral monopoly after contracting and investing in specific assets has occurred. The antitrust laws are aimed at protecting consumers from high prices and reduced quantities caused by exploitation of a monopoly position. However, the appropriation of quasirents does not lead to a price increase to consumers. There are transfers between firms, but they do not effect market prices since quasirents, being returns on sunk investments, do not effect prices.

Moreover, the main difference in this context between antitrust and contract law is that antitrust has multiplied (trebled) damages. The theory behind such damages is that some violations will not be detected, so that penalties must be greater to provide sufficient deterrence. However, contractual opportunism occurs in situations where there are two well-defined parties dealing with each other. Therefore, such opportunism would be detected with a very high probability, so that there would be no benefit from multiplied damage payments. There might well be some costs in terms of overdeterrence, since multiplied damage payments might cause some firms to maintain contracts even when it would be more efficient to break them. For these reasons, it is appropriate to treat situations involving exploitable quasirents under contract law, but not under antitrust law.

CHANGING CIRCUMSTANCES

Where there are specific assets, changes in circumstances may lead to changes in efficient modes of cooperation and gains from cooperation may increase or decrease. When there is no integration, and assets are separately owned, owners of assets will then bargain over responses to changed circumstances and over the division of gains or losses from these same circumstances. Such bargaining is costly and time consuming. While the parties are bargaining, they are not acting, and therefore money is being lost. An additional benefit of organization within a firm is the power of the owner or manager to order terms; of transactions when the best set of terms changes, and therefore avoid this loss from bargaining. The right to control the use of assets within a firm is sometimes called theresidual decision right.In situations where change is likely or where delays from bargaining are costly, organization within a firm is more desirable.

ASSET SPECIFICITY

In the example of the landfill and the hauling company, the relation between the two assets takes the form it does because of a geographic linkage between the two assets. This is a common form of asset specificity. Another example of site-created specificity would be the relationship between a coal-burning electrical generation plant located at the mouth of a coal mine and the mine itself. We would expect common ownership. Similarly, Alcoa aluminum bought many bauxite mines because it often paid to build smelters at the site of the mine, presenting obvious holdup problems if mines and smelters were owned by different parties. Common ownership of steel production and steel milling can be explained in the same way. It is desirable to have steel milling located near steel production because of the savings in reheating costs, but this common siting could create holdup problems without common ownership.

But geographic linkage is not the only form that asset specificity can take. Another example is physical asset specificity. An investment may be undertaken in a product which is useful to only one customer. The traditional example is the early relation between General Motors and Fisher Body. Originally, GM contracted with Fisher for auto bodies. But Fisher realized that it would be subject to a holdup problem once it build a plant to make bodies for GM as the car maker would then be in a position to exploit the quasirents associated with the plant. Therefore, Fisher contractually required GM to buyallof its bodies from Fisher, thus removing GM's potential holdup power. But this, on the other hand, created power for Fisher to hold up GM. GM tried to protect itself contractually by fixing prices and other terms of the transaction. Demand for bodies grew faster than anticipated, leading to lower costs for Fisher, and Fisher's profits were greater than GM had anticipated. GM was unhappy with a situation giving Fisher excess profits based on an erroneous market forecast. Ultimately, GM and Fisher merged.

INFORMATION PROBLEMS

A seller may exploit a buyer by selling a lower quality product than was expected or agreed upon. This will increase the profits of the seller at the expense of the buyer. The seller can know more about quality than can the buyer, which is an example ofasymmetric information,a general problem which occurs frequently. Vertical integration can eliminate this incentive to exploit. The general point is this: The more difficult it is to measure quality of an intermediate good, the more likely it is for vertical integration to be justified. Additionally, ownership of goods should change hands at that point in the production process where quality monitoring and measurement is least costly.

In deciding whether to make or buy an input, these principles are relevant and should be considered by a manager. He should be more willing to buy, rather than make, a good whose quality is easy to measure. Additionally, where there is an option, a manager should structure the purchase decision so that he is buying at a point where measurement is relatively easy. For example, a firm which needs dyed cotton cloth for an input may find it most efficient to buy greige goods (undyed cotton cloth) and perform the dying itself because costs of measuring quality of undyed material should be lower than costs of measuring quality of dyed material.

ALTERNATIVE STRATEGIES

While the presence of exploitable quasirents is necessary for vertical integration to be worthwhile, it is not sufficient. The presence of potential holdup problems does not by itself guarantee that such integration will be beneficial. Vertical integration reduces possibilities for cost control as compared with use of markets, so the cost of the holdup problems must be greater than the loss in efficient cost savings for such integration to pay.

Returning to our original example, if it would cost the hauling company $.50 per ton more than the landfill company to operate the landfill, then vertical integration would not pay. If the hauling company wanted to buy the landfill, it would have to pay a price for the landfill based on current conditions, assuming the lowest cost of running the landfill. If the hauling company would be less efficient than the current operator in running the landfill, then it would suffer a loss on its purchase. If this loss were more than the exploitable quasirent (calculated at $.40 per ton), then it would be cheaper to be exploited than to buy the landfill. The landfill would undertake a similar analysis of its costs of running the hauling company.

It is at this point that the traditional make-or-buy analysis begins. The traditional accounting or cost analysis assumes the data which has been discussed. That is, this accounting analysis examines the lowest cost alternative for a given set of costs for each action. But the interesting managerial questions are those related to the data inputs into the accounting process. Unless the correct structure is placed on the problem, including the correct measures for costs of being victimized by opportunistic behavior, the answer from the accountant will be wrong.

When there are specific assets and exploitable quasirents but it does not pay to make the product internally, a difficult problem arises for the firm. In these circumstances it becomes important to protect oneself from exploitation. There are many contractual methods of providing such protection, which will be discussed in the next chapter.

OTHER EXPLANATIONS

Vertical integration is sometimes explained by technological factors. The example above of steel milling and steel production is often cited as a technological explanation for common ownership. It is said that both operations must be under such ownership because the technology requires it.

There are true benefits from having steel production and milling in one location. These are technological, and are based on the savings in reheating costs associated with rolling steel when it is still hot. But a technological linkage by itself does not require vertical integration and common ownership. Technology does not dictate ownership. For example, it is quite possible to think of a steel mill buying hot steel ingots from a furnace located in the same building through an arm's-length transaction, with no common ownership. There is a well-defined market price for steel of a certain grade, and one can easily think of a contract specifying a mechanism for determining the transaction price based on market prices, with the savings from avoiding extra reheating split between the parties.

But no steel mill would put itself in the position of being dependent on a single furnace for its supply of steel. No furnace would put itself in the position of being forced to sell all of its output to a single mill. Common ownership is required by the transaction structure of the deal, not by the technology.

It is sometimes argued that vertical integration is justified because a firm needs a secure source of supply and can only establish such a secure source by integrating. For example as a professor, writer, and consultant, I cannot function without a secure source of paper clips, paper and ink for my printer. Nonetheless, I do not find it worthwhile buying paper mills, paper clip factories, and ink manufacturers. I do not worry about a secure source of supply because these items are readily available on the market.

When would someone worry? Only if the supplier is in a position to hold up the buyer. If a small office supply store in an isolated town induced me to move there by promising a secure source of paper and other supplies, I might be subject to a holdup problem after I had moved. The quasirents which could be exploited would be the costs of moving. In this case, I might want to use some method (although obviously not vertical integration) to guarantee my supplies. But in the normal course, where no holdup possibility exists, vertical integration is not needed to guarantee security. It should not be surprising that those cases where vertical integration is justified on the basis of securing supplies are cases where asset specificity and other conditions for potential holdup prevail.

Some argue that firms must sometimes buy inputs because they lack the capital needed to acquire facilities for internal production. It is very unlikely that this will be the case if tine cost analysis is done correctly. When holdup problems occur, this means that the aggregate value of both firms under common ownership is greater than their value under separate ownership. A correctly performed accounting analysis will demonstrate this. This analysis should be a powerful tool to bring to the capital markets. The landfill is worth $3.5 million (the present value at 10 percent of $460,000 per year) more to the trucking company than it is worth to the current owner. If the current value of the landfill is $40 million, then its value to the truck company is $43.5 million, a substantial increment. The hauling company can pay a premium over the landfill's current value and still come out ahead, and it is difficult to imagine that there is no lender who would be willing to finance such a profitable transaction.

WHO SHOULD OWN AN ASSET?

So far, we have focused on the decision as to the location of an activity: Should an input be produced within a firm or produced outside and bought by the firm? A related issue is the ownership of an asset: Should the firm own some asset or should the asset be owned by another firm? This decision is related to the production decision since assets are used in productive activities, but some additional insights can be gained by focusing separately on the asset ownership issue. The specific question is this: Assets A (owned by Firm A) and Assets B (owned by Firm B) are jointly used i:n production. There are three options: Firm A can own A and B; Firm B can own A and B; or Firm A can own A and firm B can own B. That is, there are two possibilities for integration (A owns everything, or B does) and one option of nonintegration.

But, first, what is meant by ownership in functional, not legal, terms? Associated with

Excerpted from Managing Business Transactions: Controlling the Cost of Coordinating, Communicating and Decision Making by Paul H. Rubin
All rights reserved by the original copyright owners. Excerpts are provided for display purposes only and may not be reproduced, reprinted or distributed without the written permission of the publisher.