Strategic Links

Vertical integration does not necessarily mean that a company's upstream plants supply inputs to its downstream operations. In fact, it is more likely that they do not.

Most people think that a firm owns a production chain for the sake of making the parts that go into its finished products. Ford's River Rouge complex in Dearborn, Michigan, which was once the largest integrated factory complex in the world, is a classic example. It contained almost every facility needed to produce Ford's vehicles, from refining raw materials to assembling automobiles. Other examples of such vertically integrated firms include power plants that control their own coal mines, oil companies that extract crude oil and sell gasoline, and clothing companies that create their own fabrics.

In this view, the aim of vertical integration is to ensure a steady supply of quality inputs that would ultimately prove cheaper than sourcing the same materials from outside the company. Although not everything can be produced internally, the transactions costs of searching and bargaining for the right inputs and making sure that a supplier sticks to the terms of a contract may induce firms to produce some of the inputs they need in order to avoid these costs, according to Nobel Prize winner Ronald Coase, a professor emeritus at the University of Chicago Law School.

But a recent study titled "Why Do Firms Own Production Chains?" by Chicago professors Ali Hortaçsu and Chad Syverson finds that although companies operate production chains, they do not necessarily use these links to facilitate the movement of inputs made by upstream plants to downstream plants that make the final goods. In fact, only a very small fraction of upstream units are dedicated to their firms' downstream operations. Thus, most companies that own vertical production links do not function in the way that is talked about in much of the academic literature and the business press.

Still, companies with vertically linked operations do exist and may choose such a structure for reasons that have nothing to do with supplying inputs. "It makes us rethink why firms expand vertically and what they are really trying to do," says Syverson. Instead of moving physical goods from one plant to another, Hortaçsu and Syverson propose that the primary purpose of vertical ownership is to mediate the efficient transfer of intangible rather than tangible inputs within firms. These intangibles include managerial skills, marketing and sales expertise, and knowledge from research and development (R&D).

This explanation is consistent with previous research that suggests that it may be harder for firms to get the exact intangible components they need outside of the company, unlike physical goods that would be relatively easier to obtain from the market. As a result, a firm may be better off transferring the skills and knowledge that it already has within the company to other business units as it expands.

Outward Bound

The authors analyzed about 30,000 upstream plants that were owned by firms that had at least one downstream unit. These upstream operations reported almost 3 million shipments of their products during the sample period.

Of these shipments, only a very small share was delivered to downstream plants in the same company. In fact, the typical upstream plant ships only 2.6 percent of the value of its output to downstream units within the firm. One-third of these plants report no internal shipments at all. Even upstream plants that could be considered relatively big suppliers of inputs to their companies' downstream operations still sell about 40 percent of their output outside the firm.

The exception to this general pattern is a small set of establishments— about 2 percent—that are dedicated to serving the downstream needs of their company. In general, the fraction of plants that ship to downstream units falls steadily as the share of internal shipments rises, but then suddenly shoots up to reflect this special group of upstream producers that deliver 100 percent of their output internally. Thus, most of the stories about vertically integrated firms are really about those 2 percent of upstream plants. "The vast majority of the remaining plants are not anywhere near that," says Syverson. "They are much closer to the other end of the scale where they are not shipping anything inside the firm."

Several checks to test the robustness of the study's results reveal the same patterns. The traditional view that firms choose to own plants to control their supply of inputs appears to be unfounded in most cases. Thus, companies must have other motivations for choosing a vertically integrated structure.

The Role of Intangible Inputs

To understand why companies own production chains, the authors first look at how plants that belong to a vertically integrated structure are different from other plants.

Hortaçsu and Syverson find that plants with vertical links are significantly more productive, larger in terms of output, and more capital-intensive than similar establishments that are not part of an integrated company. For instance, vertically linked plants produce 4.5 times more output than other plants in the same industry. Moreover, it seems that these plants were already bigger and more productive even before they were brought into the company.

The authors also observe that a vertically integrated company is usually larger than other firms that are not integrated. Thus, it is possible that the size of the firm rather than the structure of the company can explain why integrated plants tend to be larger and more productive. That is, upstream and downstream plants are typically of this "type" because large firms own bigger and better facilities, an explanation that has nothing to do with the merits of vertical integration itself. Indeed, the authors find that vertically linked plants tend to belong to big firms, and big firms tend to have better performing plants.

These results are consistent with the view that the firm is an outcome of a matching mechanism where two groups want to get together to produce something of value. A marriage market, for instance, is composed of highly skilled men and women as well as lesser skilled candidates. As one would expect, highly skilled female candidates typically match up with highly skilled male candidates.

In the same way, the study's results suggest that firms with high-quality plants search for other high-quality facilities to bring into the company. In particular, companies build up their business by acquiring plants that complement their management skills and other expertise and knowledge in marketing, sales, and R&D.

Thus, even though most integrated companies do not move physical inputs along the production chain, such intangible inputs can be efficiently transferred within a vertically integrated firm. In other words, firms own production chains because their intangible assets are a good match for the company's plants. The result is that vertically integrated production chains are found in the largest firms composed of the best performing plants.

Why Firms Expand

If this alternative explanation for vertical ownership is correct, then the distinction between an upstream plant and a downstream plant no longer applies. Managerial skills and other similar inputs are just as likely to be transferred from a firm's downstream units to its upstream units as the other way around.

Moreover, vertical integration may not be too different from a horizontal expansion of the company. When a firm expands horizontally, it typically goes into a new market that is not far from its current line of business, with the expectation that the company's strengths can be carried over to the new facilities. Such an expansion need not involve a transfer of physical inputs. Management skills and other types of expertise, on the other hand, are expected to flow to the new businesses.

Vertical expansions can operate in the same way. In this case, companies are obviously expanding into related markets because the businesses they acquire can supply or buy from each other, although that is not necessary. Whether companies expand vertically or horizontally is just an indication of the direction that the business is going, but the motivation is the same—to find markets where they can excel.

Hortaçsu and Syverson find some support for the notion that management skills in vertically integrated companies are transferred to newly acquired businesses. When a plant is bought by an integrated company, the plant's labor productivity and capital intensity go up due to an increase in physical capital from the new investment and a decline in hours worked. In fact, the fall in hours worked is mostly due to a decrease in hours worked by nonproduction workers rather than production workers.

The relative decline in nonproduction workers is consistent with replacing the plant's old managers with the firm's own set of managers who will run the newly acquired plant. Capital intensity also would be expected to rise upon integration if the company expects that its managers or other intangible inputs have the expertise to use the machines and equipment it acquires to make the new plant more productive.