The giant firm arrives with an explanation of its size. Its furnaces are more efficient, its distribution more reliable, its laboratories more productive, its purchasing more economical. What smaller enterprises have to negotiate, it can arrange; what each would have to build separately, it can share. These are serious claims. A large organization can make things possible that a small one cannot. But the account of what size permits must be kept beside the account of what size costs, and both must be distinguished from the income that size enables its owner to collect. The firm may become larger because it produces more cheaply. It may also become larger because control over more of the market is valuable to it.
Alfred Chandler gave the first explanation a historical body. In his account of modern industrial enterprise, late nineteenth-century railway and telegraph networks, alongside steamships and cables, made sustained flows of materials and goods possible. New production technologies offered potential economies of scale and scope, but realizing them required investments in distribution and management as well as plant. Rated capacity did not guarantee actual throughput. The managerial organization coordinated the flows on which low unit costs depended. Learned capabilities could then support expansion and make entry difficult for challengers. Chandler’s argument consequently joins productive organization to enduring competitive advantage; it does not make a factory’s size a sufficient explanation of its performance.
The relation can be made visible without mistaking a large invoice for an economy. Suppose two alternative facilities can make the same product, at the same quality, over a specified period. One costs thirty to establish for the period and four for each unit produced. The other costs eighty to establish and two per unit. These are hypothetical resource costs, including the costs relevant to the comparison, and both facilities can supply the quantities considered. At twenty units the first costs a hundred and ten and the second a hundred and twenty. At a hundred units the first costs four hundred and thirty and the second two hundred and eighty. The larger commitment becomes cheaper only above twenty-five units. Its capacity is an opportunity whose value depends on use.
The distinction matters before and after the investment. Before either facility is built, the fixed commitments belong in the choice: one must ask whether the expected use warrants eighty rather than thirty. After a payment has become irrecoverable, continuing production presents a different question, concerning the costs and receipts that operation can still change. A plant can then be worth operating even though its original construction was a mistake. Conversely, a plan that looked efficient at a hundred units can become expensive when orders fall to twenty. The production account needs a date as well as a scale. Otherwise the profitability of continuing an inherited plant is easily mistaken for proof that building it was wise.
Even this careful comparison concerns a facility, not yet the size of a firm. A single company can own two small plants without making either more efficient. Two independent companies can use the same large distribution service without merging. Common ownership may allow coordination or shared investments that separate contracts make difficult; it must be shown to do that work. Simply adding the balance sheets proves only that more assets now answer to the same authority. The number of establishments, the output of each establishment, and the extent of common control are different dimensions of size.
The managerial hierarchy earns its place where the work of coordinating these activities is itself consequential. A schedule that keeps an expensive machine supplied, an inventory rule that prevents a stoppage, and a distribution system that brings the output to buyers can preserve the value of a productive investment. These are acts of production in the economically relevant sense: they prevent resources from being wasted. Their contribution may be difficult to see because a successful arrangement often leaves no interruption to observe. Yet that difficulty does not entitle every additional office to credit for the plant’s entire output. The comparison is with what another arrangement could achieve, including an arrangement that buys some of these services outside.
Coase’s boundary asks whether organizing another activity costs less than the available alternatives. Organizing more transactions can save the costs of repeated market dealings, but it can also increase the costs of organization and the mistakes made within it. Activities differ in their demands; distance and dissimilarity can make additional supervision harder. A new means of communication can help the office and the market at once, so its effect on the boundary depends on which costs it lowers more. The relevant question is always comparative. A firm cannot justify an additional layer merely by showing that coordination is necessary; it must explain why that layer performs the coordination better than the available alternatives.
The same numerical example can expose the price of a bad organization. At a hundred units the larger facility saves a hundred and fifty in the costs already counted. If its arrangement requires a further hundred and sixty of administrative resources, beyond those allowed for in the first account, its total cost becomes four hundred and forty. The smaller alternative remains at four hundred and thirty. An authentic technical economy has been more than consumed by an organizational expense. If the additional expense is only forty, the total is three hundred and twenty and a gain of a hundred and ten remains. One can accept the engineering advantage and still dispute the institution built around it.
Sharing a facility across products creates another possible economy and another demand on coordination. A laboratory, a delivery network, or a specialized body of knowledge may serve several products. Sharing it can save the cost of reproducing it; sharing it can also burden the common service with incompatible demands. A useful comparison holds the outputs and their quality fixed and asks what resources separate provision and joint provision require. A company that acquires an unrelated business does not acquire an economy of scope by definition. It acquires another business. Whether the combination adds value depends on what can actually be shared, what coordination that sharing requires, and what each activity loses when it must compete for attention inside a larger whole.
The profit account introduces another distinction. Return to the hundred-unit example and suppose all units sell for five. Before any extra administrative charge, the smaller facility earns seventy and the larger earns two hundred and twenty. The increase of a hundred and fifty corresponds to the resource saving. Now hold the larger facility’s output and costs fixed but let its price rise from five to six. Profit rises by another hundred, although no additional resource has been saved. Buyers pay a hundred more for the same goods. That second gain is a transfer in this fixed-output comparison. If a higher price also excludes purchases that would have been worth producing, a separate loss can arise through the change in output. An income statement by itself does not separate these mechanisms.
This is why the question in the title cannot be answered by the firm’s survival alone. Survival can reflect the ability to use resources well, a favorable position from which to collect income, or a combination of the two. A barrier to entry may protect a difficult capability built through patient investment; it may protect an advantage that a rival could reproduce if access were available. The practical inquiry must identify the activity being protected and the alternatives being excluded. A demonstrated cost advantage does not settle how much of its benefit should reach buyers or workers. A demonstrated power to charge more does not show that every productive capability behind the firm is dispensable.
Inside the organization, the same distinction becomes a question of governance. Managers may know enough to coordinate the productive system while having reasons to favor its enlargement. Owners may possess formal rights without supervising closely. The earlier plate on the hierarchy placed this separation of ownership and control beside the rise of the large enterprise; it remains part of the inquiry into scale. The costs of monitoring and the costs of managerial discretion have to enter the comparison, alongside any gains from shared production. These costs can be worth bearing. Their existence defeats the inference that whichever arrangement has grown largest must therefore be the arrangement that wastes least.
The economy of scale must survive the cost of the organization built to obtain it.
The giant firm can be a productive achievement whose benefits depend on discipline it cannot supply merely by declaring itself efficient. Its machines must be used, its offices must add more than they consume, and its profits must be interpreted with attention to the prices and alternatives behind them. Once those questions have been asked, another remains: who should hold the authority that organizes the achievement and receive the remainder it leaves? Scale can explain a need for coordination without deciding which participants should own the coordinating institution. The cooperative takes up that unfinished choice. The argument remains open.