Here’s a puzzle: In the 1980s, U.S. automobile manufacturers imitated the lean manufacturing techniques that had been pioneered by their Japanese rivals. Around the same time, they also radically changed the way that they sourced components for their cars. Outside suppliers began to be treated more like they were internal component-producing divisions within the car company, while internal divisions were forced to function more like outside suppliers. The car companies exercised greater authority over their outside suppliers by mandating process improvements for higher quality and productivity. Outside suppliers also were required to share their private information about costs. By contrast, internal component divisions within the car company were treated more like autonomous profit centers, free to sell their components to outside companies and required to compete with outside suppliers for sales within their own corporation. Why did these changes in procurement practices happen, and why did they happen at that particular time?

Recent award-winning research by Richard Makadok and Russell Coff, associate professors of organization and management at Emory University's Goizueta Business School, answers this puzzle by developing a new theory to explain the emerging phenomenon of hybrid governance structures—new types of business relationships that infuse market-like characteristics into organizational hierarchies, or hierarchy-like characteristics into market contracts.

When manufacturers seek a source for needed components, they have traditionally faced the classic make-or-buy decision: They can either turn to the market and buy parts from outside suppliers, or they can develop expertise within their own organizational hierarchies and make components themselves. In 2009, American economist Oliver Williamson won the Nobel Prize in economics for research published in 1975 that helped explain how companies resolve the classic make-or-buy dilemma in a way that minimizes conflicts of interest. His research provided the first purely economic theory for why a firm exists and how it defines its boundaries.

But Williamson’s central dichotomy between markets and hierarchies doesn’t account for a vast range of innovative governance structures that have grown increasingly popular in the past 35 years, Makadok says. While a great many firms employ markets, hierarchies or an intermediate governance model such as a joint venture to source components, a growing number of companies are utilizing hybrid structures that infuse market-like features into hierarchies or vice versa. Such hybrid governance structures are important, Makadok says, because they offer firms a means to indirectly promote goals such as cooperation and creativity that are tough to measure, and therefore tough to motivate. In a 2009 paper, Makadok and Coff developed a model that describes when and why various hybrids work, and gives insights into how various hybrid governance structures can provide indirect inducements for tasks that otherwise could not be directly incentivized by any motivational lever because of measurement problems. They argue that firms can identify situations where such a “leverless” task synergistically makes it easier to perform “levered” tasks, which can easily be measured and incentivized. A carefully designed hybrid governance structure can induce just the right amount of effort on both the levered and leverless tasks alike.

“Our model is based on the classic problem of how to influence behaviors that lack motivational levers,” they write. “It addresses why a rational principal might counterintuitively choose a hybrid governance form that rewards one behavior in hopes of achieving another. Our model highlights how incentive patterns that appear superficially to be dysfunctional can nevertheless be quite effective when the right pattern of cross-task synergies is present.”

For example, Makadok notes that it is often argued that tying teachers' compensation to their students performance on standardized tests is counterproductive, because it motivates teachers to "teach to the test"—drilling students excessively on the relatively simple and rote skills that can be measured on a standardized test, while ignoring, or at least excessively de-emphasizing, higher-level skills such as creativity and judgment that cannot easily be measured on standardized tests. “I have no reason to doubt that this is a real problem,” says Makadok, “but it is only a problem if no synergies exist between developing the two types of skills. As a thought experiment, imagine what would happen if developing higher-level skills like creativity and judgment made students better at lower-level skills, or at least made it easier for students to learn those lower-level skills. Then those synergies, if they existed, could eliminate, or at least mitigate, the ‘teaching to the test’ incentive problem, because teachers would know that their efforts to teach the higher-level skills of judgment and creativity would, at least indirectly, boost student performance on standardized tests.”

The paper by Makadok and Coff was published in the Academy of Management Review, an influential research journal in the field of management, whose editorial board recently accorded it the “Best Paper Award” for 2009. The paper’s title—“Both Market and Hierarchy: An Incentive-System Theory of Hybrid Governance Forms”—is a play on the title of Williamson’s 1975 book, Markets and Hierarchies: Analysis and Antitrust Implications. It’s a twist that’s best explained by a phenomenon that hadn’t yet happened when Williamson’s book was written—the puzzle mentioned earlier about car companies changing their procurement practices.

A key aspect of the lean manufacturing system that U.S. car companies adopted in the 1980s is what’s called “design for manufacturability,” in which product design is simplified and streamlined so that products can be assembled more easily. This, in turn, helps manufacturers cut costs and improve quality. Plus, it lets companies move to a “just-in-time” inventory management system that further boosts efficiency. What’s intriguing about this shift to lean manufacturing, Makadok says, is that it had important implications for the governance structures of car companies. In the case of external suppliers, for example, assemblers previously had used arm’s length transactions to buy parts. But with lean manufacturing, assemblers needed to rely on these suppliers to do more of the work on designing parts, so that they could be designed in ways that enabled efficiency gains. What’s more, suppliers needed to be flexible in meeting an assembler’s demand for parts, which fluctuated with consumer demand for final products. External suppliers, in other words, no longer sold just parts—they also sold a service.

“Previously, if I was an assembler, I was just buying the parts themselves, and so the supplier’s most important assets were the physical plant and equipment,” Makadok explains. “But now I’m also buying the supplier’s expertise about design and manufacturability, as well as flexibility on the part of suppliers to adjust their outputs according to shifts for demand in my product. So now, what becomes the most important assets are the supplier’s design and engineering skills and reputation for flexibility.”

With this shift, car companies found they could more easily achieve their goals with hybrid governance structures that strayed from the classic continuum between markets and hierarchies. Traditionally, parts suppliers were either completely internal or external to firms. With internal suppliers, assemblers owned the physical assets of the business units and exercised considerable authority over decisions about how and when to manufacture parts. In exchange, internal parts divisions were given only weak financial incentives for production. With external suppliers, assemblers had no ownership stake and gave them considerable authority in manufacturing decisions. In exchange, these outside suppliers also took on the financial risks and rewards that come with the market.

But with lean manufacturing, car companies turned to hybrid structures that scrambled how markets and hierarchies typically handled the key organizational variables of authority, ownership and incentives. For internal suppliers, companies moved to a hybrid that Makadok calls an “autonomous profit center,” in which business units were forced to compete with outside suppliers for the assembler’s business. In exchange, the units were given elements of entrepreneurial risk—including the chance to sell parts to other assembly companies—and the opportunity to reap the financial rewards of their success. In the case of external suppliers, car companies brought these outside firms closer to their own organizational hierarchies through a hybrid structure that Makadok calls “quasi-integration.” In these cases, the degree of risk in suppliers’ compensation was significantly reduced. Suppliers often received long-term contracts and guaranteed minimum purchase agreements that cut their downside risk. In exchange, assemblers asserted more authority over how suppliers manufactured parts and obtained much more visibility into suppliers’ cost structures. Makadok and Coff quote an observation made by a financial analyst that summed up the relationship this way: “Toyota knows its suppliers’ cost structure better than its suppliers do. You may never get rich being a supplier to Toyota, but you will never ever go out of business.”

The shift to autonomous profit centers and quasi-integration in the automobile industry created a new continuum of governance structures, where the poles were no longer defined in terms of markets and hierarchies. Rather, the make-or-buy decision shifted to one where assemblers selected hybrids based on whether the supplier of a given part needed to work cooperatively with other suppliers. In some cases, for example, assemblers needed suppliers to cooperate with other firms so the designs of their respective subsystems wouldn’t conflict in ways that might impede manufacturability. Cooperation is one of those “leverless” tasks, so how could it be motivated? Quasi-integration was the governance structure that worked in these cases because the supplier’s incentive to maintain key assets—namely, a reputation for reliable performance and intelligently designed parts— also provided an incentive for cooperation. Assemblers, in turn, no longer needed to provide strong productivity incentives to these suppliers. Where cooperation among suppliers was not needed to make certain parts, car companies delegated the tasks to internal units. Rather than structure the units along the lines of traditional hierarchies, assemblers provided new incentives for the divisions to increase productivity and sell to other outside assemblers. In doing so, assemblers structured the units much as professional service firms organize their senior partners as autonomous profit centers.

Makadok points out that the switch to lean manufacturing spurred the switch to these new hybrid structures because asset maintenance became synergistic with other tasks. Traditionally, external suppliers struggled with a counter-synergistic relationship between maintaining the asset of their physical plants and producing parts for the assembler. The reason? Suppliers can’t make parts when the equipment is shut down for maintenance, and can’t maintain equipment when it’s being used to produce output. But in the lean manufacturing context, asset maintenance became synergistic with production because intangible and reputational assets, rather than physical equipment, became key to success. And those intangible assets can be maintained more easily when they are being used. As the study authors write in the paper: “Meeting output expectations, helping other agents and serving other principals all may help build a reputation for reliable performance.” Moreover, a supplier’s design skills are also enhanced by using them, in learning-by-doing fashion.

While Makadok and Coff devote a significant portion of their paper to discussing the two hybrid structures, the autonomous profit center and quasi-integration, adopted by car companies in the 1980s, those aren’t the only governance options described by their model. In fact, the authors chart out the parameters for six hybrids that vary depending on whether certain features are more like a hierarchical organization or more like an independent entity in the market. In particular, the model creates a sort of “governance cube” where hybrid structures occupy distinct zones within the cube, and the three axes of the cube represent the three key dimensions of governance: (1) ownership of assets; (2) rewards for production; and (3) authority to determine how the work gets done and whether other outside activities, such as moonlighting, are allowed. The traditional hierarchical organization sits at one extreme corner of the cube, since the firm owns the assets, offers weak rewards for production and has strong authority over operations. The traditional market organization sits at the opposite corner of the cube, since the supplier owns its assets, has strong production incentives and retains full authority for making its own decisions about what other activities it will pursue. The six governance structures outlined by Makadok and Coff are situated in distinct zones within the cube, which graphically shows how hybrids mix either two parts hierarchy structure and one part market structure, or two parts market and one part hierarchy.

A “franchising” arrangement, for example, resembles a market structure in two key respects: franchisees are independent companies; they also reap most of the operating rewards, meaning they have strong productivity incentives. Yet the franchisor provides the business model and ensures that it is strictly followed by the franchisee—an authority relationship that’s akin to how principals control divisions within organizational hierarchies. This particular hybrid, Makadok notes, accounts for more than 13 percent of U.S. gross domestic product, and over 33 percent of retail sales take place through franchised businesses. At the other end of the spectrum, a structure the study authors describe as “empowerment” features two parts hierarchy and one part market. In organizations that emphasize human capital such as research universities, for example, the principal retains ownership of the assets and provides weak financial incentives for production. But the structure of the research university is a hybrid because faculty members engage in self-governance that’s more akin to independent contractors, rather than relying on formal authority found in most hierarchical organizations.

The two researchers use the term “piece-rate employment” to describe another hybrid that’s more hierarchical than market-oriented. Here, principals own assets and retain strong authority but gives workers strong productivity incentives. An example is a sales force where workers are paid primarily through large commissions. For the final hybrid, Makadok and Coff use the term “consortium.” The example here is an open-source consortium in software, where members are market-like entities with distinct asset ownership and no central authority over members’ actions. Yet given that members tend to be volunteers, there’s also an element of hierarchical organizations—weak productivity incentives.

“What our paper does is really focus on better explaining nonstandard governance forms, which we call hybrids,” Makadok says. “But this paper is not simply a taxonomy. The paper is also about explaining when each of these governance forms will be observed.”

The “governance cube” and the ideas about governance structures that it summarizes may seem abstract, Makadok acknowledges. But it’s an attempt to resolve the very concrete problem routinely confronted by management upon recognizing that things like cooperation and creativity are central to getting a job done yet notoriously difficult to measure. It’s tough in a hierarchical or market arrangement to provide incentives for these tasks, so a hybrid governance structure can do the trick by providing indirect incentives.

“Managers are aware of when tasks cannot be directly motivated, and the notion of designing governance structures to motivate them indirectly is not hard to convey,” the authors write. “Managers would simply identify how tasks must be grouped, which ones can be directly motivated and which ones can only be encouraged indirectly. In this way, our theory may prove relatively easy to apply in an actual organizational context.”