Tài liệu Microeconomics for MBAs 20 doc

Chapter 6. Reasons for Firm Incentives 9 efforts. Such a finding means that if each worker added to the group must be paid the same as all others, the cost of additional production obviously rises with the size of the working group. The finding also implies that to get a constant increase in effort with the additional workers, all workers must be given greater incentive to hold to their previous level of effort. 12 Optimum Size Firms How large should a firm be? Contrary to what might be thought, the answer depends on more than “economies of scale” technically specified. Technology determines what might be possible, but it doesn’t determine what will happen. And what happens depends on policies that minimize shirking and maximize the use of the technology by workers. This means that scale economies depend as much or more on what happens within any given firm as they do on what is technologically possible. The size of the firm obviously depends on the extent to which owners must incur greater monitoring costs as they lose control with increases in the size of the firm and additional layers of hierarchy (a point well developed by Oliver Williamson in his classic article written more than thirty years ago 13 ). However, the size of the firm also depends on the cost of using the market. Management information Professors Vijay Gurbaxani and Seungjin Whang have devised a graphical means of illustrating the “optimal firm size” as the consequence of two forces: “internal coordinating costs” and “external coordinating costs.” 14 As a firm expands, its internal coordinating costs are likely to increase. This is because the firm’s hierarchical pyramid will likely become larger with more and more decisions made at the top by managers who are further and further removed from the local information available to workers at the bottom of the pyramid. There is a need to process information up and down the pyramid. When the information goes up, there are unavoidable problems and costs: costs of communication, costs of miscommunication, and opportunity costs associated with delays in communication, all of which can lead to suboptimal decisions. These “decision information costs” become progressively greater as the decision rights are moved up the pyramid. Attempts to rectify the decision costs by delegating decision making to the lower ranks may help, but this can – and will -- also introduce another form of costs -- which, you will recall, we previously have called agency costs. These include the cost of monitoring (managers 12 Workers can also reason that if the residual from their added effort goes to the firm owners, they can possibly garner some of the residual by collusively (by explicit or tacit means) restricting their effort and hiking their rate of pay, which means that the incentive system must seek to undermine such collusive agreement. For a discussion of these points see, Felix R. FitzRoy and Kornelius Kraft, “Cooperation, Productivity, and Profit Sharing,” Quarterly Journal of Economics (February 1987), pp. 23-35. 13 Oliver E. Williamson, “Hierarchical Control and Optimum Size Firms,” Journal of Political Economy, vol. 75 (no. 2, 1967), pp. 123-138. 14 Vijay Gurbaxani and Seungjin Whang, “The Impact of Information Systems on Organizations and Markets,” Communication of the ACM, January 1991, pp. 59-73. Chapter 6. Reasons for Firm Incentives 10 actually watching employees as they work or checking their production) and bonding (workers providing assurance that the tasks or services will be done as the agreement requires), and the loss of the residual gains (or profits) through worker shirking, which we covered earlier. The basic problem managers face is one of balancing the decision information costs with agency costs and finding that location for decision rights that minimizes the two forms of costs. From this perspective, where the decision rights are located will depend heavily on the amount of information flow per unit of time. When upward flow of information is high, the decision rights will tend to be located toward the floor of the firm, mainly because the costs of suboptimal decisions by having the decision making done high up the hierarchy will be high. The firm, in other words, can afford to tolerate agency costs because the costs of avoiding the them, via centralized decisions, can be higher. Nevertheless, as the firm expands, we should expect that the internal coordinating costs along with the cost of operations will increase. The upward sloping line in Figure 6.1 depicts this relationship. But internal costs are not all that matter to a firm contemplating an expansion. It must also consider the cost of the market, or what Gurbaxani and Whang call “external coordination costs.” If the firm remains “small” and buys many of its parts, supplies, and services (such as accounting, legal, and advertising services) from outside venders, then it must cover a number of what we have called “transaction costs.” These include the costs of transportation, inventory holding, communication, contract writing, and contract enforcing. However, as the firm expands in size, then these transaction costs should be expected to diminish. After all, a larger firm seeks to supplant market transactions. The downward sloping line in Figure 6.1A depicts this inverse relationship between firm size and transaction costs. Again, how large should a firm be? If a firm vertically integrates, it will engage in fewer market transactions, lowering its transaction costs. It can also benefit from economies of scale, the technical kind mentioned earlier. However, in the process of expanding, it will confront growing internal coordination costs, or all of the problems of trying to move information up the decision making chain, getting the “right” decisions, and then preventing people from exploiting their decision making authority to their own advantage. The firm should stop expanding in scale and scope when the total of the two types of costs -- external and internal coordinating costs -- are minimized. This minimum can be shown graphically by summing the two curves in Figure 6.1A to obtain the U-shaped curve in Figure 6.1B. The optimal (or most efficient/cost-effective) firm size is at the bottom of the U. This way of thinking about firm size would have only limited interest if it did not lend itself to a couple of additional observations, which permit thinking about the location, shape, and changes in the curve. First, the exact location of the bottom will, of course, vary for different firms in different industries. Different firms have different capacities to coordinate activities Chapter 6. Reasons for Firm Incentives 11 through markets and hierarchies. Second, firm size will also vary according to the changing abilities of firms to coordinate activities internally and externally. ___________________________________ Figure 6.1A and 6.1B External and Internal Coordinating Costs As the firm expands, the internal coordinating costs increase as the external coordinating costs fall. The optimum firm size is determined by summing these two cost structures, which is done in the bottom half of the figure. _______________________________________ __ A firm that is efficient at processing information will be larger, everything else equal, than one that isn’t so able. If a firm is able to improve the efficiency of its upward information flow and reduce the number of wrong decisions, then the upward sloping curve in Figure 6.1A will move down and to the right, causing the sum of the two curves in the bottom panel of the figure to move to the right, for a greater optimal size firm. If the costs of using markets go down, the firm size can be expected to decline, not because the firm has become less efficient internally (it may have become more efficient), but because markets are now relatively more cost effective. Again, from this perspective, the size of the firm changes for reasons other than those related to the technology of actual production. It depends on the ability of managers to squeeze out the scale economies that are possible from their workers. Of course, knowing that the owners will always worry that their manager-agents will exploit their positions for their own benefit at the expense of the owners, managers will want to “bond” themselves against exploitation of their positions. (And we don’t use the term “bond” in the modern pop-psychology sense of developing warm and fuzzy relationships; rather, we use it in the same sense that is common when accused criminals post a bond, or give some assurance that they will appear in court if released from jail.) That is to say, managers have an interest in letting the owners know that they, the managers, will suffer some loss when exploitation occurs. Chapter 6. Reasons for Firm Incentives 12 Devices such as audits of the company are clearly in the interest of stockholders. But they are also in the interest of managers by reducing the scope for managerial misdeeds, thus increasing the market value of the company – and the value of its managers. By buying their companies’ stock, manager-agents can also bond themselves, assuring stockholders that they will incur at least some losses from agency costs. To the extent manager-agents can bond themselves convincingly, the firm can grow from expanded sources of external investment funds. By bonding themselves, manager-agents can demand higher compensation. Firms can be expected to expand and contract with reductions and increases in the costs of developing effective managerial bonds. 15 Changes in Organizational Costs Finally, we can observe that the size of the firm can be expected to change with changes in the relative costs of organizing a given set of activities by way of markets and hierarchies. For example, suppose that the costs of engaging in market transactions are lowered, meaning markets become relatively more economical vis a vis firms. Entrepreneurs should be expected to organize more of their activities through markets, fewer through firms. Then, those firms that more fully exploit markets, and rely less on internal directions, should be able to increase the payments provided workers and other resources that they buy through markets, collectively leaving fewer resources to expand their market share relative to those firms that make less use of markets. Accordingly, firms should be expected to downsize, to use a popular expression. An old, well-worn, and widely appreciated explanation for downsizing is that modern technology has enabled firms to produce more with less. Personal computers, with their ever- escalating power, have enabled firms to lay off workers (or hire fewer workers). Banks no longer need as many tellers, given the advent of the ATMs. One not-so-widely-appreciated explanation is that markets have become cheaper, which means that firms have less incentive to use hierarchical structures and more incentive to use markets. And one good reason firms have found markets relatively more attractive is the rapidly developing computer and communication technology, which has reduced the costs of entrepreneurs operating in markets. The new technology has lowered the costs of locating suitable trading partners and suppliers, as well as negotiating, consummating, and monitoring market-based deals (and the contracts that go with them). In terms of Figure 6.1, the downward sloping transaction costs curve has dropped down and to the left, causing the bottom of the U to move leftward. “Outsourcing” became a management buzzword in the 1980s because the growing efficiency of markets, through technology, made it economical. Outsourcing continued apace in the 1990s. Of 26 major companies surveyed, 86 percent said they outsourced some activity in 15 See Michael C. Jensen and William H. Meckling, “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics, vol. 3 (October 1976), pp. 325-328. Chapter 6. Reasons for Firm Incentives 13 1995, up from 58 percent who gave the same response in 1992, with the budding outsourcing industry generating $100 billion in annual revenues by 1996. 16 For all practical purposes, airlines now outsource the acquisition of their reservations through independent contractors called travel agents, given that more than 70 percent of all airline reservations are now taken by such agents, working through computerized markets, not through the hierarchical structures within the airlines. Modern technology has also improved the monitoring of employees, reducing agency costs, which has been a force for the expansion of firms. This is because firms have been able to use the technology to garner more of the gains from economies of scale and scope. The optical scanners at grocery store checkout counters are valuable because they can speed up the flow of customers through the checkout counters, but they can also be used for other purposes, such as inventory control and restocking. Each sale is immediately transmitted to warehouse computers that determine the daily shipments to stores. The scanners can also be used to monitor the work of the clerks, a factor that can diminish agency costs and increase the size of the firm. (We are told that even “Employee of the Month Awards” are made based on reports from scanners.) Books on Tape, a firm that rents audio versions of books, tracks its production of tapes by way of scanners not so much to reward and punish workers, but to be able to identify problem areas. In terms of Figure 6.1, the upward sloping curve moves down and to the right, while the U-shaped curve in the lower panel moves to the right. Frito-Lay has issued its sales people hand scanners in part to increase the reliability of the flow of information back to company distribution centers, but also to track the work of the sales people. The company can obtain reports on when each employee starts and stops work, the time spent on trips between stores, and the number of returns. The sales people can be asked to account for more of their time and activities while they are on the job. Obviously, we have not covered the full spectrum of explanations for the rich variety of sizes of firms that exists in the “real world” of business. We have also left the net impact of technology somewhat up in the air, given that it is pressing some firms to expand and others to downsize. The reason is simple: technology is having a multitude of impacts that can be exploited in different ways by firms in different situations. Prisoners’ Dilemma Problems, Again The discussion to this point reduces to a relatively simple message: Firms exist to bring about cost savings, and they generate the cost savings through cooperation. However, cooperation is not always and everywhere “natural”; people have an incentive to “cheat,” or not do what they are supposed to do or have agreed to do. This may be the case because of powerful incentives to toward noncooperation built-in to many business environments. 16 As reported by John A. Byrne, “Has Outsourcing Gone Too Far?” Business Week, April 1, 1996, p. 27. Chapter 6. Reasons for Firm Incentives 14 An illustration of the tendency toward noncooperative behavior, despite the general advantage from cooperation, is a classic so-called “conditional-sum game” known as the prisoners’ dilemma (which we have already introduced without formally calling them by their proper name). 17 This is a dilemma, commonly found in business, that takes its name from a particular situation involving the decision two prisoners have to make on whether or not to confess to a crime they committed. But the dilemma can also be applied whenever two or more people find themselves in a situation where the best decision from the perspective of each leads to the worst outcome from the perspective of all. Consider a situation in which the police have two people in custody who are known to be guilty of a serious crime, but who, in the absence of a confession by one of them, can be convicted only of a relatively minor crime. How can the police (humanely) encourage the needed confession? One effective approach is to separate the two prisoners and present each with the same set of choices and consequences. Each is told that if one confesses to the serious crime and the other does not, then the one who confesses receives a light sentence of one year, while the one who does not confess receives the maximum sentence of fifteen years. If they both confess, then both receive the standard sentence of ten years. And if both refuse to confess, then each is sentenced to two years for the minor crime. The choices and consequences facing the prisoners are presented in the “payoff” matrix in Table 6.1, where the first number in each parenthesis is the sentence in years received by prisoner A, and the second number is the sentence received by prisoner B: Table 6.1 Prisoners’ Dilemma B Don’t Confess Confess Don’t Confess (2 2) (15 1) A Confess (1 15) (10 10) From the perspective of both prisoners the best outcome occurs if neither one confesses (they serve a total of four years), and the worst outcome occurs if both confess (they serve a total of 20 years). In other words, if both prisoners cooperate with each other by keeping their mouths shut, they will both be far better off than if they act noncooperatively with each other by confessing. However, from the perspective of each prisoner the best choice is the noncooperative one of confession. Consider the situation from prisoner A’s vantage point. If A believes that B will refuse to confess, then he receives two years in prison if he also refuses to confess, but only one year if 17 “Conditional-sum games” are games in which the value available to the participants is dependent how the game is played. Chapter 6. Reasons for Firm Incentives 15 he does confess. His best choice is to confess. On the other hand, if A believes that B will confess, then he receives fifteen years in prison if he does not confess and only ten years if he does. Again, his best choice is to confess. No matter what A believes B will do, it is in A’s best interest to confess. And the incentives are exactly the same for B. So while it is rational from their individual perspectives for both A and B to make the noncooperative choice, the result is the worst possible outcome from their collective perspective. When, as in our example, only two people are in a prisoners’ dilemma setting, it is quite possible for them to avoid the worst outcome by choosing the cooperative option of not confessing. The two prisoners may be good friends and have genuine regard for the well being of each other, in which case each will feel confident that the other will not betray him with a confession, and will refuse to betray his friend. But if the number of prisoners grows and becomes quite large, then it becomes much less likely that any one of them can reasonably trust everyone else to keep quiet. This means that as the number grows it becomes increasingly irrational for any one of them to keep quiet. Overcoming the Large-Numbers Prisoners’ Dilemma Problems Overcoming a large-number prisoners’ dilemma by motivating cooperative behavior is obviously difficult, but not impossible. The best hope for those who are in a prisoners’ dilemma situation is to agree ahead of time to certain rules, restrictions, or arrangements that will punish those who choose the noncooperative option. For example, those who are jointly engaging in criminal activity will see advantages in forming gangs whose members are committed to punishing noncooperative behavior. The gang members who are confronted with the above prisoners’ dilemma will seriously consider the possibility that the shorter sentence received for confessing will hasten the time when a far more harsh punishment for “squealing” on a fellow gang member is imposed by the gang. The problem illustrated by the prisoners’ dilemma is a very general one that is encountered in many different guises, most of which have nothing to do with prisoners. Excessive pollution, for example, can be described as a prisoners’ dilemma in which citizens – meaning, typically, a very large number of people -- would be better off collectively if everyone polluted less, yet, from the perspective of each individual the greatest payoff comes from continuing to engage in polluting activities no matter what others are expected to do. As another example, while there may be wide agreement that we would be better off with less government spending, each interest group is better off lobbying for more government spending on its favorite program. People are tempted by the noncooperative solution in polluting and lobbying because they benefit individually and only have limited and costly ways of ensuring that others resist the noncooperative solution. Chapter 6. Reasons for Firm Incentives 16 Many areas of business are fertile grounds for the conditional-sum game situations represented by the prisoners’ dilemma. A number of examples of business-related prisoners’ dilemmas will be discussed in some detail in subsequent chapters, and an important task of managers is to identify and resolve these dilemmas as they arise both within the firm and with suppliers and customers of the firm. Indeed, we see “management” as concerned with finding resolutions of prisoners’ dilemmas. Good managers constantly seek to remind members of the firm of the benefits of cooperation and of the costs that can be imposed on people who insist on taking the noncooperative course. Consider, for example, the issue of corporate travel, which is a major business expense -- estimated at over $130 billion in 1994 (the latest available data at this writing). 18 If a business were able to economize on travel costs, it would realize significant gains. And much of this gain would be captured by the firms’ traveling employees who, if they were able to travel at less cost, would earn higher incomes as their net value to the firm increased. So all the traveling employees in a firm could be better off if they all cut back on unnecessary travel expenses. But the employees are in a prisoners’ dilemma with respect to reducing travel costs because each recognizes that he or she is personally better off by flying first class, staying at hotels with multiple stars, and dining at elegant restaurants (behaving noncooperatively), than making the least expensive travel plans (behaving cooperatively) regardless of what the other employees do. Each individual employee would be best off if all other employees economized, which would allow her salary to be higher as she continued to take luxury trips. But if the others also make the more expensive travel arrangements, she would be foolish not to do so herself since her sacrifice would not noticeably increase her salary. Airlines have recognized the “games” people play with their bosses and other workers, and have played along by making the travel game more rewarding to business travelers, more costly to the travelers’ firms, and more profitable to the airlines – all through their “frequent-flier” programs. Of course, you can bet managers are more than incidentally concerned about the use of frequent-flier programs by employees. When American Airlines initiated its AAdvantage frequent-flier program in 1981, the company was intent on staving off the fierce price competition that had broken out among established and new airlines after fares and routes were deregulated in 1978. As other writers have noted, American was seeking to enhance “customer loyalty” by offering their best, most regular customers free or reduced-price flights after they built up their mileage accounts. Greater customer loyalty can mean that customers are less responsive to price increases, which could translate into actual higher prices. 19 At the same time, there is more to the issue than “customer loyalty.” American figured that it could benefit from the obvious prisoners’ dilemma their customers, especially business 18 As reported in Jonathan Dahl, “Many Bypass the New Rules of the Road,” The Wall Street Journal, September 29, 1994, p. B1. 19 For a discussion of frequent-flier programs as a means of enhancing customer loyalty, see Adam M. Brandenburger and Barry J. Nalebuff, Co-opetition (New York: Currency/Doubleday, 1996), pp. 132-158. Chapter 6. Reasons for Firm Incentives 17 travelers, are in. By setting up the frequent-flier program, American (and all other airlines that followed suit) increased the individual payoff to business travelers for noncooperative behavior. American did this under its frequent-flier program by allowing travelers to benefit from more free flights and first-class upgrades by choosing more expensive, and often less direct, flights. They encouraged business people to act opportunistically, to use their discretion for their own benefit at the expense of everyone else in their firms. For example, a business traveler who is on the verge of having enough miles in his American account to qualify for elite status (additional upgrades of travel perks) might choose a more expensive American flight over a comparable Southwest Airline flight just to get additional AAdvantage miles. The company would in effect, pick up the cost of the traveler’s vacation flight. Business travelers are also encouraged to book their flights later than they could, which requires paying full fare, so they can use their frequent-flier upgrades to first class (these upgrades are typically not allowed with discount tickets). Or business people will take circuitous routes to their destinations to qualify for more frequent-flier miles than could be gotten from a direct trip. The prisoners’ dilemma problem for workers and their companies has, of course, prompted as a host of other non-airline firms -- rental car companies, hotels, and restaurants – to begin granting frequent-flier miles with selected airlines for travel services people buy with them, encouraging once again higher-than-necessary travel costs. The company incurs the cost of the added miles plus the lost time. Now, use of frequent-flier miles might actually lower worker wages (because of the added cost to their firms, which can reduce the demand for workers, and the benefit of the miles to workers, which can increase worker supply and lower wages, topics to be covered later), but, still, workers have an incentive to exploit the program. Again, they are in prisoners’ dilemma under which the cooperative strategy might be best for all, but the noncooperative strategy dominates the choice each individual faces. These problems created by frequent-flier programs are not trivial for many businesses, and we would expect the bigger the firm, the greater the problem (given the greater opportunity for opportunistic behavior in large firms). Thirty percent of business travelers working for Mitsubishi Electronics America wait until the last few days before booking their flights, according to corporate travel manager John Fazio. Fazio adds, “We have people who need to travel at the last minute, but it’s not 30 percent.” 20 Corporate travel managers complain that the frequent-flier programs have resulted in excessive air fares (a problem for 87 percent of the firms surveyed), wasted employee time (a problem for 68 percent of the surveyed firms), use of more expensive hotels (a problems for 67 percent of the surveyed firms), and unnecessary travel (a problem for 59 percent of the surveyed firms). 21 The corporate travel managers 20 See Dahl, Ibid., p. B1. 21 As reported by Frederick J. Stephenson and Richard J. Fox, “Corporate Strategies for Frequent-Flier Programs,” Transportation Journal, vol. 32, no. 1, (Fall 1992), pp. 38-50. The 1991 survey included 506 corporate members of the National Business Travel Association who did not work for airlines. Chapter 6. Reasons for Firm Incentives 18 interviewed felt that the frequent-flier programs resulted in an average “waste” of about 8 percent of all of their travel expenditures. 22 Frequent-flier programs put business travelers in a game situation that benefits the airlines at the expense of business travelers and their firms by encouraging noncooperative behavior. Recognizing this game, and the noncooperative incentives built into it, is important for managers who are trying to cut travel costs. And in the effort to cut these costs, managers are also in a game with the airlines, which respond to cost cutting measures with new wrinkles designed to intensify the prisoners’ dilemma faced by business travelers. For example, USAir announced plans to provide a Business Select class (featuring roomier seats and better meals) for those business travelers who pay full fare for their coach tickets. 23 Of course, when all airlines have frequent-flier programs, the problems for firms may be compounded by the fact that all airlines have more “loyal” customer bases and all are less likely to cut prices (another topic to be addressed later in greater detail). The Moral Sense Much our analysis in the “Manager’s Corner” sections of the chapters to this book will be grounded, as it has already, in the principal/agent problem, or the tendency of underlings to pursue their own private goals at the expense of the goals of the firm and its owners. We do that for a simple reason: We want to understand how employees might behave in order that managers can draw up policies and incentives that can protect the firm and its owners from agency costs. We do not by any means wish to suggest that people are not, in the slightest degree, driven by an innate sense of duty or obligation to do that which they are supposed to do as a employee in a team or firm. On the contrary, people do seem to have a built-in tendency to cooperate -- to a degree. UCLA business professor James Q. Wilson has shown, with reference to casual observation and to a host of psychological experiments, that most people do have a “moral sense,” which can show up in their willingness to forgo individual advantage (or opportunities to shirk) for the good of the group, which can be a firm. 24 Moreover a variety of factors -- including considerations of equity and fairness -- influence people’s willingness to cooperate. As organizational behaviorists have shown, “culture” has an impact on the extent of cooperation. People from “collectivistic” societies, like China, may be more inclined to cooperate than people from “individualistic” societies, like the United States. 25 Training in “group values” can affect the extent of cooperation. Experiments 22 Ibid., p. 41. 23 Ibid., p. 43. 24 James Q. Wilson, The Moral Sense (New York: Free Press, 1993). 25 See P. Christopher Earley, “Social Loafing and Collectivism: A Comparison of the United States and the People's Republic of China,” Administrative Science Quarterly, vol. 34, n4 (Dec, 1989), pp. 565-582. . local information available to workers at the bottom of the pyramid. There is a need to process information up and down the pyramid. When the information. balancing the decision information costs with agency costs and finding that location for decision rights that minimizes the two forms of costs. From this