October 19, 2009
Competition and Organizational Efficiency
Competition and Organizational Efficiency-Becker
Oliver Williamson's influential contributions to the theory of firms were the stimulus for our discussion topic this week of the analysis of organizations. Posner gives an excellent discussion of various factors that determine organizational structure and efficiency, such as conflicts between principles, like stockholders, and agents, like employees and managers, the ease of communicating information and knowledge from the bottom to the top of the organization, and the number of "layers" in the command structure. I will concentrate my comments on the environment that organizations face, and especially on the degree of competition they have to deal with. One of the most compelling observations from highly competitive environments is that many different organizational structures sometimes survive in the same industry. For example, in the retail grocery sector, large "warehouse" types of stores exist alongside small specialized grocery stores. Chains that own many supermarkets, such as Safeway and Whole Foods, compete against small mom and pop stores with few paid employees.
George Stigler argued many years ago in a classic article ("The Economies of Scale", reprinted in his collection of essays called The Organization of Industry) that different types of firms that survive in the competition for profits in very competitive environments must be of rather equal efficiency at producing profits. A corollary is that if a competitive industry were trending over time toward a narrower set of organizational types, this would imply that these types must have become relatively more efficient as the economic and political environments changed over time.
The fact that small supermarkets and large warehouse markets survive the tough competitive pressure of the retail grocery market suggest that both types must be of about the same efficiency in their respective niches of the grocery sector, although the trend seems to be toward larger supermarkets. That steel mills are much larger than textile factories suggest that economies of scale in steel production must be sufficiently larger than the scale economies in the production of textiles to overcome the larger number of command layers and other inefficiencies of a larger production scale in steel but not in textiles.
Also of relevance to understanding the efficiency of different organizational types is that very different types of firms survive in different countries, often even when they are in the same or similar industries. For example, Japan and South Korea (occupied by Japan for about 40 years in the 20th century) have large conglomerates that are active in many different industries, such as Korea's SK company whose products range from an oil refinery to cell phones, whereas Taiwan tends toward smaller firms that are more concentrated in particular sectors (although Taiwan was also occupied by Japan).
Both the inter country and within country evidence indicate that no single organizational form is always the most profitable even in a particular sector of the economy. Different combinations of scale economies, principle-agent problems, compensation practices, thickness of the span of control, and many other variables highlighted in the organizational literature often produce outcomes that are about equally efficient and profitable. The outcome of strong competition is the only really decisive way to determine which are the possibly quite different but about equally efficient combinations of all these different variables.
The major difficulty in evaluating many governmental organizations is that they often do not face such strong competition and they have no simple measure of success, such as profits. These two factors make it difficult to use Stigler's survival test. To take Posner's example, can the criminal catching activities of say the FBI be efficiently combined with a terrorist deterrent function? If this were a competitive industry with many different organizations and good observable measures of success, one could then look at whether such combinations compete successfully in the longer run against more specialized agencies. Lacking either much competition or such measures of success, one has to rely in good part on the insights of analyses of these issues. Similarly, the organization and efficiency of armies is only rarely tested against competition on the battlefield. When so tested, losing armies often try to reorganize so that they can look more like the successful armies, although generals are often accused of reorganizing to fight the last war.
Perhaps then it is best to try to create competition among governmental agencies, such as both the CIA and FBI trying to deter terrorism, and then provide greater resources to the more successful agency. This, however, runs into the difficulty that agencies may withhold information from each other in order to gain an advantage in such competition.
The Economics of Organizations--Posner
Oliver Williamson, an economist who won half a Nobel prize last week, has made important contributions to a field of economics that is not as well known as it should be: "organization economics." This is a field, closely related to a branch of sociology called organization theory, to which pioneering contributions were made by Alfred Chandler, Herbert Simon, and Ronald Coase, as well as Williamson; more recent contributors of note include Jacques Crémer, Bengt Holstrom, Luis Garicano, Canice Prendergast, Jean Tirole, and others. I have used organization economics in my academic work on the structure of our national intelligence system; Garicano and I have published an organization-economic study of the FBI's domestic intelligence branch in the Journal of Economic Perspectives, and I have written a review essay on organization economics for a forthcoming issue of the Journal of Institutional Economics.
Oddly, an interest in organizations is a latecomer to economics, even though most economic activity is conducted through organizations. The standard economic model is of trade between individuals, or firms assumed to behave as individuals. For many purposes the model, despite its extreme simplification, is adequate. If one wants to know how cigarette producers will respond to a rise in cigarette taxes, it is enough to assume unrealistically that a cigarette producer is one person rather than a complex organization. But for other questions the assumption is inadequate--most obviously if the question is why some business firms have steeply hierarchical structures and others rather flat ("M-shaped"--"M" standing for multidivisional) ones (this distinction has been a particular emphasis in Williamson's work). Or why compensation practices within firms (or government agencies) take the form they do. Or--most fundamentally--why there are firms at all--why all economic activity isn't carried on by contracts among individuals. Ronald Coase asked that question in a paper entitled "The Nature of the Firm," published in the 1930s. His answer was that a producer has a choice between contracting with independent contractors for the output of the various inputs into this production of the finished product, and contracting with individual workers--employees--not for their output but for the right to direct their work--and that the employer would choose between forms of contract--the contract with the independent producers or the employment contract--on the basis of which was more efficient, given the nature of his business.
Neither form of organizing production is perfect. The arms' length contract form requires detailed specifications that create inflexibility. The command form--the employer directing the work of employees rather than contracting for their output--creates the well-known principal-agent problem (the problem economists call "agency costs")--the employee is supposed to be working to maximize the firm's profits, but what he wants to maximize is his own utility, so the employer has a control problem.
The modern literature emphasizes the principal-agent problem but also moves beyond it by emphasizing another aspect of control within an organization: the creation, transmission, processing, coordination, and use of information. Because the span of supervision by one person is limited, the more employees a firm has, the more supervisors it requires; and the more supervisors it has, the more supervisors of supervisors it requires because the span of control is limited at every tier of the hierarchy. So as an organization expands, the layers of supervisors multiply, and the consequences ared delay in executing orders, loss of information, attenuation of the directions emanating from the top, and in short a weakening of control and coherence. The larger the organization, moreover, the more difficult it will be to correlate the work of a particular employee with the value of the organization's output, and so the employee's incentives will fall further out of alignment with those of the firm. A partial alternative to hierarchy is to decentralize the organization in imitation of the market, by delegating authority to division heads and requiring them to compete with one another for allocations of capital from central management. That is the essence of the "M-form" of corporate organization ("M" standing for multidivisional).
Organization economics emphasizes the variety of agency costs that flourish in complex organizations, such as "influence activities," by which agents try to influence the decisions of their principals, for example by flattery, by being a "yes man" and not "rocking the boat," by doing personal favors, by making alliances with coworkers, by jockeying for promotion, and by hoarding information to make oneself indispensable and reduce the output of one's competitors in the organization.
The challenge to organizations is to generate cooperation without use of the price system, since the employer does not buy the output of his employees. Instead organizations rely on common norms, understandings, customs, and perspectives that substitute for explicit contracting and thus enable cooperation on dimensions of performance that cannot be prescribed by formal directives. This set of informal binding elements (the organization's "culture") includes codes and other shared specific human capital that facilitates communication and coordination among agents. Unfortunately, an organizational culture that is optimal in its current environment may become suboptimal when the environment changes, yet adaptation to the new environment may be difficult because once information channels and other organizing elements are created, an investment has been sunk that will constrain the organization's reaction to a new environment. Change is especially hard because an organization's culture is diffused throughout the organization rather than concentrated in one place (an employment manual, for example) where it could be changed at a stroke. The result is organizational conservatism or inertia, and explains why innovations tend to come from new firms rather than from existing ones.
An important aspect of organizational culture, one that I have emphasized in my academic work, is the awkwardness of combining different cultures in the same organization. An example is the combination of criminal-investigation and security-intelligence functions in the FBI. The former lend themselves to what are called "high-powered" incentives, which are systems of compensation and promotion that are based on objective performance criteria. In the case of criminal investigation these are number of arrests weighted by convictions and sentence. Intelligence work does not lend itself to such performance criteria, because the effect of surveillance and other intelligence activities in preventing terrorism or subversion is usually very difficult to assess. Hence motivation takes the form of creating a "high commitment" environment in which the organization's leaders try to elicit good performance by getting staff to internalize the organization's goals. The problem is that the absence of objective criteria of performance opens the door to "influence activities" by which members of the organization jockey for advancement.
If both types of task are combined in the same organization--those that can be directed by high-powered incentives and those that require high commitment as their motivator, the best employees will tend to gravitate toward the first type of task because they will be confident that they will do well if their performance is judged according to objective criteria. They will be much less certain how well they will do in a job in which influence activities play a large role in determining success.
The problem of culture clash in an organization is further illustrated by the financial collapse of last year. Banks had traditionally been conservative organizations emphasizing risk avoidance, modest compensation, gradual promotion, and secure tenure. When in the deregulation era they were permitted to expand into riskier and (therefore) more lucrative forms of financial intermediation, they attracted a different kind of employee--smarter, more willing to take career as well as financial risks, more independent, and demanding higher pay. Because they were generating more profits for the bank, their influence grew and placed pressure on the traditional bankers to take more risks in order to hold their own in the struggle to control the organization. So one proposal for preventing a recurrence of the financial crisis, since the crisis was due in part to highly risky lending by banks, is to restore the separation codified in the Glass-Steagall Act of conventional banking from high-risk forms of financial intermediation.
The financial collapse illustrated another facet of organization economics as well. The banking industry expanded very rapidly in the low-interest-rate environment created by Greenspan's monetary policy in the early 2000s, and the expansion took the form largely of the expansion of existing firms rather than the creation of new ones. When an organization expands rapidly, there is a danger of loss of control over subordinate employees. The danger in the case of the banking industry's expansion was increased by the fact that many of the new hires consisted of young risk takers whose attitudes and skills were often quite different from those of the higher tiers of management. Senior managers had difficulty in assessing and limiting the highly risky deals engineered by the young hot shots.