Paying respect

Forthcoming in Journal of Economic Perspectives Tore Ellingsen is the Ragnar Söderberg Professor of Economics and Magnus Johannesson is Professor of Economics, both at the Stockholm School of Economics, Stockholm, Sweden. Their e-mail addresses are < JEL classification: D01, D23, D82, Z13 Keywords: Esteem, Incentives, Recognition, Personnel economics. Why do people work? Economic theory generally, and the principal-agent model specifically, emphasize the role of material incentives. The standard assumption is that people work hard only if they receive monetary compensation for doing so. However, a substantial body of evidence contradicts the standard economic model. Many academics, for example, work diligently year after year for a nearly fixed real salary, continuing to work hard as they approach retirement, when financial incentives are usually absent. We will argue that while economists have been right to focus on incentives, they have been wrong to focus so exclusively on material incentives. While workers appreciate monetary rewards, they also get utility from what (they believe that) others think about them. Thus, employers can pay their workers with a combination of monetary rewards and respect. Employers can help to fulfill employees' desire for respect in at least four ways, each of which sheds some light on observed practices. First, attention from an employer can enhance the employee's sense of being respected for good performance. Second, a manager or the employer can increase the value of respect by proving to be a worthy audience; for example, hard-working employees appreciate the respect of hard-working managers, and idealistic workers appreciate the respect of an idealistic employer. Third, the employer can influence the norms of respectability—that is, what traits gain respect in a certain workplace. Fourth, employers can affect what information about an employee is available to others, and therefore the respect or esteem that the employee receives from others. We begin by laying out a body of evidence that, taken as a whole, makes a strong case that respect matters in the workplace, above and beyond material rewards. We discuss evidence that workers respond to attention, symbolic rewards and trust – and even that material incentives in some cases lead to less effort. Building on the theoretical work in Ellingsen and Johannesson (2007a), we then argue that many of these observations can be captured in a standard principal-agent model, once the principals and the agents are assumed to care about respect or esteem as well as money. Our approach is an example of behavioral agency theory, which includes two main lines of research. The social preference approach studies the impact on contracts of other-regarding motives such as altruism, competitiveness, reciprocity, or fairness (Akerlof, 1982; Rotemberg, 1994; Auriol and Renault, 2001; Englmaier and Wambach, 2006; Fehr, Klein, and Schmidt, 2007). The social esteem approach focuses instead on the effects of self-regarding motives such as pride and shame. The two approaches overlap to some extent, because altruism and fairness are praiseworthy and selfishness is shameful, but in this paper we focus more on social esteem than on social preferences. Of course, the basic insight that workers care about receiving respect from others is an old one, and not only in psychology; for references to the economics and philosophy literatures on esteem, see Frey (1997) and especially Brennan and Pettit (2004). However, the task is just getting underway of modeling 1 It is likely that a human desire for social esteem—that is, to feel proud and to avoid shame in the context of broader society, evolved millennia ago, presumably because approval has been associated with material and sexual benefits. For anthropological evidence, see Fessler (2004) and the references therein. For some formal evolutionary arguments, see Fershtman and Weiss (1998a,b). However, the consequences of social esteem preferences can be studied systematically how respect is paid by employers, how it is received by employees, how it
complements monetary payments, and how these interactions manifest themselves in labor
markets.
Evidence that Respect Matters in the Workplace

A variety of worker surveys suggest that workers want more than just monetary compensation, and in particular want a sort of appreciation and recognition from their employers that convey “respect.” For example, “full appreciation for of work done” is the only job reward factor that consistently ranks among the top two motivators for U.S. workers throughout the post-World War II period (Wiley, 1997). Among ten factors, interesting work was considered somewhat more important in the 1980s, and good wages were at the top of the list in the 1990s, but neither interesting work nor good wages were among the top four reward factors in more than one decade out of three. Likewise, Elsdon (2002) reports that lack of recognition or appreciation is a major reason why people leave organizations, second only to lack of career development opportunities. Financial motives occupy fourth place. In this section, we describe evidence that employees value and respond positively to signs of respect, such as symbolic awards, attention and trust. Conversely, workers sometimes
respond negatively to what they perceive as signs of disrespect, including intrusive
managerial control and even, in some cases, monetary incentives.
Symbolic Rewards

Popular management books revel in symbolic awards. Examples include Bob Nelson's 1001 Ways to Reward Employees, Dianna Podmoroff's 365 Ways to Reward and Motivate Your Employees Every Day: With Little or No Money, and Cindy Ventrice's Make Their Day! Employee Recognition that Works. The huge market for these books and their authors’ speeches and consultancy service strongly suggest that the economic theory of motivation is incomplete. Increasingly, the messages of the popular management books are backed up by systematic evidence. Many field experiments have investigated the extent to which attention can be reliably used as a reward. Markham, Scott, and McKee (2002) provide an apt illustration. Over a year, studying 1100 workers in four cut-and-sew garment factories in the mid-Atlantic area, the authors compare the effect of a public recognition program with three types of controls. Inspired by the advice of practitioners, detailed in the rich if non-academic management literature, the researchers developed a public recognition program with three main ingredients: personal attention, public celebration, and mementos. All employees with perfect attendance during an entire month had their names posted with a gold star for that without a theory of their origin. Throughout, we therefore leave open the issue of why people want to make a favorable impression. month, and employees with no more than two absences during a quarter received a personal card notifying and congratulating them. At the end of the year, a public plant-wide meeting recognized good and perfect attendance. Employees with a perfect attendance record received engraved gold necklaces (females) and penknives (males), whereas employees with a good attendance record received similar mementos in silver. The program reduced absenteeism by about 40 percent and was popular with the workers. The three control treatments produced only insignificant changes in absenteeism. The analysis of such behavioral management interventions has been a lively area of management science for more than two decades. By now, more than 100 separate scientific studies have considered rewards of money, feedback, and social recognition. Although many studies consider only one reinforcer, and social recognition is less frequently studied than money and feedback, a survey of 72 behavioral management studies by Stajkovic and Luthans (2003) concludes that social recognition has a positive effect on behavior both alone and in combination with other reinforcers. Borrowing a term from cultural anthropology, the company can hypercognize certain worker traits—that is, it can determine to some extent what traits are highly recognized within this organization—and thereby shape the organizational culture. When the company gives prizes for attendance, as in the study by Markham, Scott and McKee (2002), it is utilizing its power to define the sources of esteem. For most workers, bonuses and stock ownership plans yield very little economic return to their own effort, and are thus primarily symbolic (Baron and Kreps, 1999, p. 264). However, by rewarding group effort rather than individual effort, the company conveys the message that cooperation is esteemed more highly than competition. Indeed, we believe that if such weak incentives were awarded on an individual basis, they would be interpreted as disrespectful and could even backfire. Individual material incentives only work well if they are sufficiently strong or if the employer can use other means to counteract the impression of disrespect; otherwise, symbolic rewards can be preferable. Attention

Some of the earliest work on human relations in the workplace started as an examination of how physical work conditions affected employees, but then ended up studying the relationship between workers' productivity and the amount of attention being paid to them. The so-called Hawthorne experiments were carried out between 1924 and 1932 at the Hawthorne Works, a plant of the Western Electric Company near Chicago, Illinois. The experiments initially set out to investigate the effect of working conditions – such as lighting 2 Some writers suggest that economic theories themselves are part of society's culture and therefore affect behavior. According to Pfeffer (1994, Chapter 4) and Ghoshal (2005), economics affect the values and expectations of managers, and managers' expectations tend to be self-fulfilling. If employers treat employees as unscrupulous opportunists, opportunism flourishes. And if business schools teachers assume that there is a conflict of interest between owners and managers, then MBA-trained managers become more prone to pursuing their own goals. If so, there is all the more reason to consider carefully the assumptions of principal-agent theory . – on productivity. Surprisingly, both more and less light appeared to cause higher worker effort. When workers were interviewed about their motives, a common response was that they were pleased to receive so much attention and therefore tried to do their best. The classic descriptions and interpretations of the Hawthorne experiments are Roethliesberger and Dickson (1939) and Mayo (1945); for a compact account, see Scott (2003, pages 61-62). According to Scott, “[I]t is only a slight exaggeration to suggest that the academic field of industrial sociology first saw the light of day at the Hawthorne plant." The Hawthorne experiments themselves have been controversial (Jones, 1992), and subsequent empirical studies attempting to identify Hawthorne effects in the field have reached mixed results. Perhaps it is just too difficult to separate the pure effect of increased employer attention to workers from the promise of a reward or the threat of punishment. If workers think that they are being monitored more closely because the employer is looking for ways to reduce worker rents, increased monitoring may even have negative effects (Enzle and Anderson, 1993). Evidence from a variety of laboratory and field experiments settings, in both psychology and economics, can also be informative about the basic psychological mechanisms. However, this evidence can only be extrapolated to the workplace with caution. Many studies attempt to isolate a pure effect of observation on behavior in a context without monetary rewards, which is obviously different than a workplace that offers a combination of the two. Also, the observation that occurs in these studies is often by a party who is more of a peer than an authority figure, and many people react differently to peers than to authorities. With those caveats in mind, Zajonc (1965) is prominent in the psychological literature for arguing that being observed affects performance. Zajonc offers evidence that the presence of others has a beneficial effect in the case of well-learned tasks, but a detrimental effect in the case of novel tasks. Since the former case is probably the most relevant for the workplace setting, this evidence is congruent with our argument. Relatedly, Wicklund and Duval (1971) argue that the presence of spectators increases the performer's awareness of the gap between attained performance and ideal performance, producing an unpleasant feeling. The performer exerts effort in order to reduce the unpleasantness. We refer to Seta and Seta (1995) for a more extensive discussion of the sizeable subsequent literature. Economics laboratory experiments offer some insights into the effects of being observed, too. Falk and Ichino (2006) study experimentally the behavior of German high school students conducting a simple task without a monetary incentive. They find that especially the worst performers work harder in the presence of others than by themselves. In the dictator game, one subject (the dictator) decides how to share a sum of money with another subject (the recipient). The dictator is free to choose any division. A dictator who is less anonymous – that is, more observed—tends to give more (Hoffman et al., 1994). In field studies of charitable donations, increased publicity also causes higher donations (Harbaugh 1998; Soetevent, 2005). 3 Social esteem effects seem to be at work even under conditions of great anonymity, perhaps not even requiring that the actor and the audience are present at the same time. One obvious example is that many people leave a tip at restaurants in a foreign place, although they will never visit the place again and nobody has recognized them. Many people care about the judgment of others, even when they are completely anonymous. For an intriguing piece of experimental evidence, see Dana, Cain, and Dawes (2006). While these findings represent evidence of positive audience effects, they do not prove that people consciously seek respect or esteem. Indeed, some of the sensitivity to attention
appears to be instinctive. Haley and Fessler (2005) document that a set of painted eyes
induces more generous behavior in a dictator game experiment. Bateson, Nettle and Roberts
(2006) similarly document that, compared to a control image, an image of a pair of eyes
almost tripled the contributions to an honesty box used to collect money for drinks in a
university coffee room.
Trust
The literature on human relations has long suggested that leadership style can be important. A series of empirical studies reported by Stogdill and Coons (1957) found that high levels of supervisor trust, friendship and respect was associated with better performance, at least if combined with organizational skills. Later research failed to confirm the optimistic conjecture that a trustful supervisory style is generally more profitable than a more controlling style (Hollander and Julian, 1969). However, from the point of view of conventional agency theory, it is problematic why trust should ever induce trustworthiness. The management literature is replete with stories of how trust has been beneficial. We have been impressed with the case of Svenska Handelsbanken, one of Sweden's largest banks, under the leadership of Jan Wallander. In 1970, soon after taking over as chief executive officer, Wallander abandoned budgeting and delegated most lending decisions all the way down to the personnel at each local office, arguing that trust promotes initiative and trustworthiness. The strategy helped turn around the troubled bank. Since then, the bank has remained highly profitable. For an autobiographical account, see Wallander (2003). A more frequently cited example is David Packard's (1995, p.135) account of how he became convinced that General Electric made a mistake to distrust their employees. “GE was especially zealous about guarding its tool and parts bins to make sure employees didn't steal anything.” As a result, stealing became almost like a sport. When Packard started Hewlett and Packard, “[T]he GE memories were still strong and I determined that our parts bins and storerooms should always be open.” Of course, trust will not always be rewarded. Some researchers report that a reduction in monitoring induces some workers to shirk more, and that it increases average shirking (Nagin et al., 2002). However, even in this work, a significant fraction of the workers do not take advantage of the increased incentive to cheat. Thus, the key issue with trusting behavior by managers is not whether the manager’s trust is sometimes abused, but whether the benefits of trust are outweighed by the costs. Economic experiments have helped to document and clarify the ways in which trust elicits trustworthy behavior. Early laboratory experiments considered behavior in the so-called trust game, originally devised by Berg, Dickhaut, and McCabe (1995). In this two-player game, the players start out with endowments e1 and e2 respectively. Player 1 then 4 According to incomplete contract models, principals may delegate their power as a commitment to assure the agent of credible protection against future intervention by the principal. Such delegation is not a sign that the principal trusts the agent, but a device to enable the agent to trust the principal. decides how much money x to transfer to Player 2. Player 2 receives ax, where the parameter a is set by the experimenter and is typically 2 or 3. Finally, Player 2 decides how much money y to transfer back. In the end, Player 1 thus gets a payoff of e1-x+y dollars, whereas Player 2 gets e2+a x -y dollars. Note that x is a measure of Player 1's trust, whereas y is a measure of Player 2's trustworthiness. The original study establishes a robust positive correlation between x and y. A number of studies elucidate the underlying mechanisms by varying the sets of actions that are available to the players. Comparing two different choice sets for player 1,
McCabe, Rigdon and Smith (2003) find that voluntary trust induces Player 2 to behave more
trustworthily than involuntary trust. Fehr and Rockenback (2003) and Fehr and List (2004)
conduct trust game experiments in which Player 1 has a choice whether or not to punish an
opponent who returns less than their requested back-transfer. Falk and Kosfeld (2006) devise
a related game in which Player 1 does not make any transfer at all, but can restrict the options
of Player 2. All these experiments find that intentions matter, not just actions. When Player 2
recognizes that Player 1 actively chose to act in a trusting manner, that act is typically
rewarded.
Possible Negative Effects of Incentives

In some specific settings, monetary incentives actually seem to produce a negative effect on the desired outcome. For example, Titmuss (1970) famously argued that the supply of human blood decreased in the United States in the 1960s because altruistic blood donors resented the use of material payments. In a recent field experiment, Mellström and Johannesson (2006) find that prospective female blood donors indeed respond negatively to a monetary incentive, whereas the behavior of potential male donors is largely unaffected. Gneezy and Rustichini (2000a) present experimental evidence that a small material incentive can induce less effort than under no material incentive. In their first experiment, a small positive piece rate for problem-solving reduces the number of solved problem compared to a zero pay benchmark. Likewise, in their second experiment, a small positive piece rate for collecting money to charity entails lower collection effort than does a zero pay benchmark. Gneezy and Rustichini (2000b) conduct a field experiment in which daycare centers impose a monetary fine on parents who collect their kids too late. The imposition of the fine results in an immediate and significant increase in late collection; apparently when parents can pay to pick their children up later, they have less hesitation about doing so. Finally, Lepper, Green and Nisbett (1973) find in an experiment that children who have previously been awarded for a task are less willing to conduct the task later. This finding echoes a previous experiment by Deci (1971), who found that people's desire to carry out a task diminishes if they have previously received monetary rewards for it, even if the reward was a surprise and did not constitute an incentive. For a survey of the vast subsequent literature, see Deci, Koestner, and Ryan (1999). 5 Several other experiments document the effects of intentions. For example, Charness (2004) shows that intentions matter in gift exchange games. Extrapolating this kind of evidence to a typical workplace must be done with caution. But it does suggest that the display of certain desirable traits, such as concern for other people, will not be especially encouraged by monetary payments.
Worthy Managers and Organizations

Employees often value nonmonetary rewards, but sometimes they are an occasion for
mockery. Employees value some kinds of attention, but not if that attention is perceived as intrusive or exploitive. Employees value trust, but some will take advantage of it. Employees value monetary rewards, but in certain settings such rewards can be counterproductive. One key determinant of whether employer actions are interpreted as showing respect or disrespect is the workers’ perception of the character of their employers. Some workers evidently care intensely about the public image of their top managers or their organization. In fall 2004, the news media in Sweden reported that the chief executive officer of the Red Cross in Sweden, Christer Zettergren, had chosen a Jaguar as his company car. Neither Zettergren nor other top management had anticipated that the brand of the car would be of any relevance to Red Cross workers or their trade unions. The car cost less than many other cars that would be viewed as acceptable. Yet, the uproar was about the brand and not the cost. The conflict ended immediately when Zettergren changed the car to a less “fancy” brand. Many Red Cross workers are idealists, and part of the value of working for Red Cross is the satisfaction of an idealistic image, both in their own minds and in the eyes of others.
Indeed, employees in the non-profit sector earn substantially less than they would do in a
comparative job in a for-profit company (Frank, 2003). From this perspective, Zettergren's
behavior was problematic for two reasons. First, if outsiders believe that working for Red
Cross is like any other job, fully compatible with a selfish and materialistic orientation, and
perhaps even as well-paid, Red Cross workers would lose social esteem. Second, for some
workers, if top managers are unwilling to make personal sacrifices, that suggests that such
managers will not truly appreciate the sacrifices that subordinates make. Presumably, one of
the reasons for non-profit incorporation is to send a clear signal that principals are willing to
forgo economic surpluses.
Behavioral Agency Theory

The standard economic explanation for the above findings would presumably go something like this. Workers care what employers believe about them, because these beliefs translate into wage raises and promotions (Holmström, 1999). The symbolic awards are simply messages about which behaviors will yield material benefits in the future. While these explanations are partially true, the practitioners’ own interpretations, as evidenced in the popular management books, are different. Moreover, the standard model fails to explain the evidence that trust sometimes induces trustworthiness; this failure is especially clear in the controlled game experiments. Behavioral agency theory introduces a richer set of motivations into the principal- agent model. Researchers re-examine existing principal-agent models, but apply different utility functions. The social preference approach introduces concerns for others into the utility function. The social esteem approach, which we pursue here, introduces instead (or in addition) a concern for what others think. To understand the social esteem approach , consider the following relationship between a single employer and a single employee. The employer knows that the employee is inherently either mediocre or talented. The employee wants the employer to believe that he is talented. Since inherent traits are unobservable, the employer has to infer the employee’s type from the employee’s effort (or from some factor correlated with effort). Compared to the case in which the employer already knows the employee’s talent, it can be shown that talented employees will tend to work harder for an uninformed employer. The reason is that the fully informed employer’s esteem for the employee is independent of effort, whereas the uninformed employer will only believe that the employee is talented if the effort is sufficiently great. (Otherwise, if the employee is too easily convinced about the employee’s talent, even a mediocre employee would be tempted to exert effort in order to earn esteem.) In this precise sense, the uninformed employer pays respect in return for high effort. The example makes clear why respect is not the same as praise. The employer can praise the employee without the employee feeling respected. Only if praise credibly reveals the employer’s belief about the employee’s talent does it convey respect. By implication, respect is not something that can be easily traded in a market. Instead, a satisfactory formal analysis of respect requires a signaling model, albeit one in which people care about others’ beliefs directly, not merely the material consequences of those beliefs. As a general proposition, this was pointed out (albeit in other contexts) more or less simultaneously by Bernheim (1994), Glazer and Konrad (1996), and Ireland (1994). Equipped with such a signaling model, it is straightforward to illustrate the above example mathematically.
6 In doing so, theorists usually also relax the common assumption that all people have similar preferences. Much of the models’ power is caused by heterogeneity, and results often hold even if a majority of people have completely standard (purely selfish and materialistic) preferences. 7 Seminal papers along these lines are Holländer (1990) and Kandel and Lazear (1992). However, these papers emphasize workers relationship to each other rather than their relationship to the employer or boss. 8 Some economists are critical of signaling models, because they are often rejected by the data. We think that signaling models have too often been invoked in settings where they are likely to be inappropriate, such as financial markets. Since financial markets tend to aggregate information quite efficiently, it is unlikely that companies should have to engage in costly signaling to convey their information to the stock market. Our setting is very different, in that there is no aggregation of information, and the actor cares about the opinions of everyone in the audience. 9 Let e denote effort, let c1e denote a talented employee’s effort cost, and let c2e denote mediocre employee’s effort cost. Let p(e) denote the employee’s belief that the employee is talented – the employer’s respect for the employee. Finally, let rp denote the employee’s “utility of respect.” Suppose there are no material rewards for effort. Under the assumptions that 0<c1<c2 and r>0, the only outcome to satisfy the so-called Intuitive Criterion is for a With this background, it should be apparent why employee attention can have a positive effect on employee effort. Suppose that the employee’s effort (or a factor correlated with effort) is being observed with probability q. If q is zero, effort is incapable of affecting social esteem. As q increases, the effect of effort on the employee’s expected esteem increases. To create a role for symbolic rewards, we may assume either that the employee’s concern for social esteem is affected by such symbols, or that the agent has many different characteristics, each of which may potentially be praiseworthy. For example, both talent and punctuality are praiseworthy, but a companywide campaign for punctuality can make respect in this domain relatively more important. Modeling corporate culture in this fashion would seem largely consistent with the analysis of Hofstede (1980). Extending the model to account for the positive effects of trust is more demanding. In Ellingsen and Johannesson (2007a) we assume that both principals and agents differ in their degree of selfishness. Moreover, we make the key assumption that everyone dislikes appearing selfish in the eyes of others, especially if these others are themselves not selfish. We are then able to identify situations in which only relatively unselfish principals trust and only relatively unselfish agents behave in a trustworthy manner. In other situations, all types of principals trust, while only relatively unselfish agents behave in a trustworthy manner. Finally, in yet other situations, no type of principal will trust. The logic works roughly as follows. An employer taking an action that appears to be selfish, like intrusive monitoring of workers to push for greater effort, undermines the workers’ incentive to signal their own unselfishness. Since workers care more about being highly esteemed by an unselfish employer, they may work harder under weaker material incentives. Generally, when control mechanisms are relatively ineffective, they will have only a modest effect in prodding selfish workers to exert somewhat higher effort, and this positive effect is outweighed by the reduction in effort by less selfish workers. As the control mechanisms get more effective, the benefit from improved control eventually dominates. At this point, all employers will benefit materially from imposing control. However, since employers also care about esteem, unselfish employers are still willing to forgo material benefits in return for more esteem. Finally, a principal-agent model along these lines allows principals to affect the esteem that their agents achieve from others in society. These others may not be able to observe the agent's performance directly, but may have access to a coarse measure of the performance that the principal provides. In corporations, awards and titles may partly serve the function of conferring a coarse signal of the esteem that the employee is due. mediocre agent to exert no effort and for a talented agent to exert effort r/c2, which is the smallest effort level that the mediocre agent would not want to mimic, even in return for maximum respect. 10 Other related models that address the puzzle that stronger material incentives can induce less effort include Bénabou and Tirole (2003, 2006), Fang and Moscarini (2005), and Sliwka (2007). In the psychology literature, Dickinson (1989) makes the related point that esteem considerations might explain the negative effects of past rewards on future interest in a task. . Directions for Future Research

Having begun to establish a formal framework, the next step is to develop further testable implications. Here are some that spring to mind. First, if employers can pay in part with respect, there should be observable monetary tradeoffs. Employees should be willing to take jobs with high social status for less monetary reward than they would otherwise require, as previously noted by Fershtman and Weiss (1993). The marginal cost of inducing additional effort from employees motivated by social status is also lower. Some employees will care more about being paid respect than others. Presumably, equilibrium sorting as well as competition between firms affect the eventual outcome, as indicated by Besley and Ghatak (2005) in their analysis of mission-oriented firms and workers. Our approach is complementary to theirs as we offer a microfoundation for their assumption that mission-oriented employees prefer mission-oriented employers.
A second possible extension of our analysis is to investigate the relationship between performance and the structure of publicly observable non-monetary awards. An analogy to grading and educational goals may be helpful here. If the purpose is to extract the maximum effort from a particular skill group, then the best grading scheme may be a pass/fail system, with the requirements for a pass set at the highest effort level that people in this group are willing to exert. If the principal cares about the effort of all types, a finer grading scheme may be optimal, but the optimal grading scheme is generically coarser than the underlying distribution of types (Zubrickas, 2007). Similarly, the ways for employers to show respect can focus on different parts of the workforce or on different skill levels, depending on the underlying main problems of production. A third issue is that the theory so far focuses on preferences about what other people think. It seems likely that people also care about what other people say. Recent experiments isolate a significant positive effect of anticipated verbal reactions on behavior in a dictator game; (Erte and Houser, 2006; Ellingsen and Johannessen, 2007b). The analogy is that anticipations of praise or disapproval by the employer will affect employee’s effort. A seemingly straightforward way to include such concerns in the model is to let the concern for social esteem depend on features of the feedback, such as emotional strength and proximity. 11 Sorting becomes particularly important if workers care about local, relative respect; see Frank (1985) and Fershtman, Hvide, and Weiss (2005). 12 To the best of our knowledge there is little previous theoretical work on optimal grading and goal-setting. The goal setting literature is largely empirical, and traditionally focuses on how to determine a single goal (although possible benefits of multiple goal levels have occasionally been discussed; Locke and Latham, 1990). Within a formal model, Costrell (1994) characterizes an optimal educational standard under the assumption of a binary (pass/fail) grading scale. In a recent contribution more closely related to ours, Dubey and Geanakoplos (2004) characterize the optimal grading scale for students who compete with each other and primarily care about relative performance. For an empirical study of how grading standards affect performance, see Figlio and Lucas (2004). Personnel economics has usually sought to analyze human resource management under the maintained assumption that employers and employees hold selfish and materialistic preferences, combined with arguments about incentives and information (as discussed by Lazear and Shaw in this issue). In the market for ideas, personnel economics has been doing well, but it is not a dominant paradigm. Business schools still largely employ psychologists and sociologists to teach human resource management. Likewise, microeconomists are conspicuously absent as authors of popular books on workforce motivation. In comparison to microeconomists’ commercial success in the fields of market design, antitrust, and corporate finance, personnel economists are lagging. Like Baron and Kreps (1999), we believe that the segmentation between economic and psychological analysis of human resource management has been detrimental. By looking only at one aspect of a problem, the analysis becomes partial, and omitted factors induce biased interpretation even of the factors under consideration. Moreover, students get a confusing picture of the field when teachers are ignorant of each others’ perspectives. Behavioral agency theory seeks to bridge the gap between the economic and the psychological approach to human resource management. If nothing else, we hope to have illustrated the general point that such bridges are desirable and feasible. Our more specific argument is that economic analysis is wrong to insist that all people always work for money only. While the simplification is sometimes innocuous, at other times it produces large discrepancies between theory and evidence. The alternative assumption that respect matters too, at least to some people in some places some of the time, helps us account for several such puzzles. Acknowledgements We thank Andrei Shleifer for his encouragement and constructive criticism, James Hines, Jeremy Stein and Timothy Taylor for their detailed comments and suggestions, and the Torsten and Ragnar Söderberg Foundation (Ellingsen) and the Swedish Research Council (Johannesson) for financial support. Parts of the paper were written while Ellingsen visited IIES at Stockholm University. Their hospitality is gratefully acknowledged.
References

Akerlof, George A. 1982. “Labor Contracts as Partial Gift Exchange.” Quarterly Journal of
Economics, 97(4): 543-569.
Auriol, Emmanuelle, and Regis Renault. 2001. “Incentive Hierarchies.” Annales d''Economie
et de Statistique, 63 (July-September): 261-282.
Baron, James N. and David M. Kreps, 1999. Strategic Human Resources: Frameworks for
General Managers, New York: John Wiley & Sons.
Bateson, Melissa, Daniel Nettle, and Gilbert Roberts. 2006. “Cues of Being Watched
Enhances Cooperation in a Real-World Setting.” Biology Letters, 2(3): 412-404.
Bénabou, Roland, and Jean Tirole. 2003. “Intrinsic and Extrinsic Motivation.” Review of
Economic Studies, 70(3): 489-520.
Bénabou, Roland, and Jean Tirole. 2006. “Incentives and Prosocial Behavior.” American
Economic Review, 96(5): 1652-1678.
Bernheim, B. Douglas. 1994. “A Theory of Conformity.” Journal of Political Economy,
102(5): 841-877.
Besley, Timothy, and Maitreesh Ghatak. 2005. “Competition and Incentives with Motivated
Agents.” American Economic Review, 95(3): 616-636.
Charness, Gary. 2004. “Attribution and Reciprocity in an Experimental Labor
Market.” Journal of Labor Economics, 22(3): 665-688.
Costrell, Robert M. 1994. “A Simple Model of Educational Standards.” American Economic
Review, 84(4): 956-971.
Dana, Jason, Daylian M. Cain, and Robyn M. Dawes. 2006. “What You Don't Know Won't
Hurt Me: Costly but Quiet Exit in Dictator Games.” Organizational Behavior and Human
Decision Processes, 100(2): 193-201.
Deci, Edward L. 1971. “The Effects of Externally Mediated Rewards on Intrinsic
Motivation.” Journal of Personality and Social Psychology, 18(1): 105-115.
Deci, Edward L., Richard Koestner, and Richard M. Ryan, 1999. “A Meta-Analytic Review
of Experiments Examining the Effects of Extrinsic Rewards on Intrinsic Motivation.”
Psychological Bulletin, 125(6): 627-668.
Dickinson, Alyce M. 1989. “The Detrimental Effects of Extrinsic Reinforcement on “Intrinsic
Motivation.” Behavior Analyst, 12(1): 1-15.
Dubey, Pradeep, and John Geanakoplos. 2004. “Grading Exams: 100, 99, . , 1 or A, B, C? Incentives in Games of Status.” Cowles Foundation Discussion Paper No. 1467. Ellingsen, Tore, and Magnus Johannesson. 2007a. “Pride and Prejudice: The Human Side of Incentive Theory.” Manuscript. Stockholm School of Economics. Ellingsen, Tore, and Magnus Johannesson. 2007b. “Verbal Feedback Induces Prosocial Behavior.” Manuscript. Stockholm School of Economics. Englmaier, Florian, and Achim Wambach. 2006. “Optimal Incentive Contracts under Inequity Aversion.” Manuscript. Harvard Business School. Enzle, Michael E., and Sharon C. Anderson. 1993. “Surveillant Intentions and Intrinsic Motivation.” Journal of Personality & Social Psychology, 64 (2): 257-266. Elsdon, Ron. 2002. Affiliation in the Workplace: Value Creation in the New Organization. Westport: Praeger. Fehr, Ernst and Bettina Rockenback. 2003. “Detrimental Effects of Sanctions on Human Altruism.” Nature, 422: 137-140. Fehr, Ernst, Alexander Klein, and Klaus M. Schmidt. 2007. “Fairness and Contract Design.” Econometrica, 75(1): 121-154. Fehr, Ernst, and John List. 2004. “The Hidden Costs and Returns of Incentives - Trust and Trustworthiness among CEOs.” Journal of the European Economic Association, 2(5): 743-771. Fershtman, Chaim, Hans K. Hvide, and Yoram Weiss. 2005. “Cultural Diversity, Status Concerns, and the Organization of Work.” Research in Labor Economics, Forthcoming. Fershtman, Chaim, and Yoram Weiss. 1993. “Social Status, Culture, and Economic Performance.” Economic Journal, 103(July): 946-959. Fershtman, Chaim, and Yoram Weiss. 1998a. “Why Do We Care What Others Think About Us?” In Avner Ben-Ner and Louis Putterman, eds., Economics, Values, and Organization, 133-150. Cambridge: Cambridge University Press. Fershtman, Chaim, and Yoram Weiss. 1998b. “Social Rewards, Externalities and Stable Preferences,” Journal of Public Economics, 70(1): 53-73. Fessler, Daniel M.T. 2004. “Shame in Two Cultures: Implications for Evolutionary Approaches.” Journal of Cognition and Culture, 4(2): 207-262. Figlio, David N., and Maurice E. Lucas. 2004. “Do High Grading Standards Affect Student Performance?” Journal of Public Economics, 88(9-10): 1815-1834. Frank, Robert H. 1985. Choosing the Right Pond: Human Behavior and the Quest for Status. New York: Oxford University Press. Frank, Robert H. 2003. What Price the Moral High Ground? Ethical Dilemmas in Competitive Environments. Princeton: Princeton University Press. Frey, Bruno. 1997. Not Just for the Money: An Economic Theory of Personal Motivation. Cheltenhem: Edward Elgar. Ghoshal, Sumantra. 2005. “Bad Management Theories are Destroying Good Good Management Practices.” Academy of Management Learning and Education 4(1): 75-91. Glazer, Amihai, and Kai A. Konrad. 1996. “A Signaling Explanation of Charity.” American Economic Review, 86(4): 1019-1028. Gneezy, Uri, and Aldo Rustichini. 2000a. “Pay Enough or Don't Pay at All.” Quarterly Journal of Economics, 115(3): 791-810. Gneezy, Uri, and Aldo Rustichini. 2000b. “A Fine Is a Price.” Journal of Legal Studies, 29(1): 1-17. Haley, Kevin J., and Daniel M.T. Fessler. 2005. “Nobody's Watching? Subtle Cues Affect Generosity in an Anonymous Economic Game.” Evolution and Human Behavior, 26(3): 245 256. Harbaugh, William T. 1998. “What Do Donations Buy? A Model of Philanthropy Based on Prestige and Warm Glow.” Journal of Public Economics, 67(2): 269-284. Hofstede, Geert. 1980. Culture's Consequences: International Differences in Work Related Values. Beverly Hills: Sage Publications. Hollander, Edwin P., and James W. Julian. 1969. “Contemporary Trends in the Analysis of Leadership Processes.” Psychological Bulletin, 71 (): 387-397. Holländer, Heinz. 1990. “A Social Exchange Approach to Voluntary Cooperation.” American Economic Review, 80(5): 1157-1167. Holmström, Bengt. 1982. “Managerial Incentive Problems: A Dynamic Perspective.” Reprinted in Review of Economic Studies, 66(1): 169-182, 1999. Ireland, Norman. 1994. “On Limiting the Market for Status Signals.” Journal of Public Economics, 53(1): 91-110. Kandel, Eugene, and Edward P. Lazear. 1992. “Peer Pressure and Partnerships.” Journal of Political Economy, 100 (4): 801-817. Lepper, Mark R., David Greene, and Richard Nisbett. 1973. “Undermining Children's Interest with Extrinsic Rewards.” Journal of Personality and Social Psychology, 28(1): 129-137. Locke, Edwin A., and Gary P. Latham. 1990. A Theory of Goal Setting and Task Performance. Englewood Cliffs, NJ: Prentice Hall. Markham, Steven E., K. Dow Scott, and Gail H. McKee. 2002. “Recognizing Good Attendance: A Longitudinal, Quasi-Experimental Field Study.” Personnel Psychology, 55(3): 639-660. McCabe, Kevin A., Mary L. Rigdon, and Vernon L. Smith. 2003. “Positive Reciprocity and Intentions in Trust Games.” Journal of Economic Behavior and Organization, 52(2): 267-275. Nelson, Bob. 1994. 1001 Ways to Reward Employees. New York: Workman Publishing. Pfeffer, Jeffrey. 1994. Competitive Advantage through People: Unleashing the Power of the Work Force. Boston: Harvard Business School Press. Podmoroff, Dianna. 2005. 365 Ways to Reward and Motivate Your Employees Every Day: With Little or No Money. Ocala: Atlantic Publishing. Rotemberg, Julio J. 1994. Human Relations in the Workplace. Journal of Political Economy, 102(4): 684-717. Scott, W. Richard. 2003. Organizations: Rational, Natural, and Open Systems. 5th Edition. Upper Saddle River: Prentice Hall. Seta, Catherine E., and John J. Seta “When Audience Presence Is Enjoyable: the Influences of Audience Awareness of Prior Success on Performance and Task Interest.” Basic and Applied Social Psychology, 16 (1-2): 95-108. Soetevent, Adriaan R. 2005. “Anonymity in Giving in a Natural Context? A Field Experiment in 30 Churches.” Journal of Public Economics, 89 (11-12): 2301-2323. Sliwka, Dirk. Forthcoming. “Trust as a Signal of a Social Norm and the Hidden Costs of Incentive Schemes.” American Economic Review. Stajkovic, Alexander D., and Fred Luthans. 2003. “Behavioral Management and Task Performance in Organizations: Conceptual Background, Meta-Analysis, and Test of Alternative Models.” Personnel Psychology, 56(1): 155-194. Stogdill, Ralph M., and Alvin E. Coons. 1957. (Eds.) Leader Behavior: Its Description and Measurement. Columbus OH: Bureau of Business Research, Ohio State University. Ventrice, Cindy. 2003. Make Their Day! Employee Recognition that Works. San Francisco: Berrett-Koehler. Wicklund, Robert A., and Shelley Duval. “Opinion Change and Performance Facilitation as a Result of Objective Self-Awareness.” Journal of Experimental Social Psychology, 7 (3?): 319-342. Zajonc, Robert B. 1965. “Social Facilitation.” Science, 149, 269-274. Zubrickas, Robertas. 2007. “Optimal Contract Design with Non-Pecuniary Rewards.” Unpublished. Stockholm School of Economics.

Source: http://www2.hhs.se/personal/ellingsen/pdf/JEPposting.pdf