The management of every corporation — particularly one that is facing a merger — struggles with how to organize its operations to maximize profits. Most large corporations are divided into business units that focus on specific products or geographic regions, while some functions are centralized at the corporate level. General Electric, for example, handles sourcing as a global activity and leaves sales, distribution and manufacturing to product business units. And at Procter & Gamble, product development, accounting, and finance are centralized, while sales, distribution, manufacturing, and procurement are managed by regional business units.
These hybrid structures always involve tradeoffs. Centralizing activities can achieve cost savings through standardization, but business-unit managers typically want to customize processes in a way that increases profits in their division. And because corporate managers and business-unit managers have different incentives and motivations, coordinated decision making and planning can be difficult.
Professor Wouter Dessein, working with Luis Garicano and Robert Gertner of the University of Chicago, analyzed how to balance these competing interests in corporations. Their findings help predict the likely success of a merger between two firms, given the firms’ organizational structures.
Consider two large car companies that are undertaking a merger. If the new management allows all the functions of the original companies to remain independent, nothing is gained from the merger. To take advantage of synergies, management may decide to centralize a function, such as product development. The new corporate manager of product development would be responsible for cost savings in the production of new cars.
This might be a good choice for the bottom line, but the corporate manager may impose excessive standardization which, in turn, can reduce revenues. To avoid a decline in sales, management may experiment with different ways of allocating authority, perhaps requiring the corporate manager to get approval for all his decisions from the business-unit managers (which can effectively bring centralized management to a standstill).
The researchers found that the best way to handle these dilemmas is to change the corporate manager’s incentives to cut costs. “If you want corporate managers to make good decisions about standardization without going overboard, you need to weaken their incentives,” Dessein says. “You need to make them care about the whole organization.”
Typically, firms benefit by giving strong incentives to managers. But if the corporate managers have very strong incentives to cut costs, Dessein explains, they will not consider the needs of the business managers. But lowering incentives has its own risks. “In the case of a stand-alone firm, would you give corporate managers a flat wage?” he says. “It depends on how effective variable pay is for managers in the company.”
In corporations that emphasize variable pay as a tool for motivation, Dessein says, both corporate and business-unit managers may be held strictly accountable for the performance of their units, but business-unit managers should then be given more authority. In this case, the corporate managers will still seek excessive standardization, but the business-unit managers will be able to block any standardization that is not to their advantage. “With this type of structure,” Dessein says, “you preserve high-powered incentives, but you will get only win-win synergies, because all the business-unit managers will have veto power.”
However, this structure runs the risk of excessive decentralization, making it difficult to manage the corporation efficiently. Therefore, the researchers propose that companies centralize those activities in which incentives are less important, such as R&D and human resources, and benefit from a standardized approach that can lead to cost savings. “For these types of functions, the incentives are naturally diffuse,” Dessein says. “If you’re in charge of product development for a whole firm with hundreds of brands, there are limits to what you can do. If you care about the profitability of each and every product, you don’t care much about any single product in particular.”
Because the goal of many mergers is often cost savings, the researchers predict that merging companies whose original structures are based on strong incentives is more likely to fail. “It’s far more difficult to realize the synergies in that case,” Dessein says. The most challenging mergers are those in which the synergies to be obtained come from many small distinct decisions. Such decisions need to made on a case-by-case basis by corporate managers whose incentive structure is such that they may impose excessive standardization. However, if all the synergies are likely to be achieved from just a few big decisions at the merger’s outset, combining the companies can be successful.
“Every firm faces these questions, but with mergers, the challenges are more visible,” Dessein says. “Merging companies often have the idea of saving costs by standardization. But they ignore the ‘organizational discount’ that should be applied in valuing the merger: the changes in the incentive structure that are required to make it work.”
Wouter Dessein is the Eli Ginzberg Professor of Finance and Economics at Columbia Business School.