Three laboratory studies and one field study show that people generally hold lay theories which contain annexations incentives bias?people predict that others are more motivated than themselves by extrinsic incentives bib security, pay) and less motivated by intrinsic incentives (learning new things). The extrinsic incentives bias can be separated from a self-serving bias and it provides an empirical counterexample to the traditional actor- observer effect in social psychology (although its theoretical explanation is similar).This kind of bias may hinder organizations from organizing because people who act as principals may use improper lay theories to offer inappropriate deals to agents. Organizations must convince their members to adopt the goals of the organization. If they do, they will meet the fundamental challenge of organizing. This fundamental challenge has attracted attention from theorists of organizations and theorists of individual motivation.
Organizational theorists have described the challenge and outlined its solution?organizations offer their members a deal: inducements in exchange for contributions (Bernard, 1938; March & Simon, 1958).Theorists of individual motivation have examined the potential content of this deal?organizations must address their members' various needs For comments on earlier drafts, the author thanks Linda Babcock, Steve Cole, Craig Fox, Tom Geologic, Dan Heath, Jill Kern, Josh Klan, Rod Kramer, Songs Laboriously, Jim March, Michael Morris, Lee Ross, Batik Westfield, John Wright, and participants in workshops at Carnegie-Mellon, Chicago, Duke, Illinois, Minnesota, and UCLA. The author thanks the Graduate School of Business, University of Chicago, and the Faux School of Business, Duke University, or research support.The author also thanks the Citron Behavioral Sciences Research Council for research support and access to the field site used in Study 4. Address correspondence and reprint requests to Chip Heath, Faux School of Business, Duke university, BOX 90120, Durham, NC 27708.
E-mail: heath@mail. Duke. Due. 25 0749-5978/99 $30. 00 Copyrighted by Academic Press All rights of reproduction in any form reserved.
26 CHIP HEATH (Moscow, 1954; Herbert, Amasser, & Snyder, 1959; Aldermen, 1972) and must combine and administer incentives so that their members are effectively motivated Broom, 1964; Needle & Lawyer, 1989; Locke & Lethal, 1990).Thus, research on organization theory has recognized the deals importance, and research on motivation has discussed the deals content. However, almost no research has discussed the deals social psychology?I. E. , how well does the person who is proposing the deal understand the person who is the deals target? "Organizations" cannot make a deal?at some point one individual (a principal) must infer what kind of deal would effectively motivate another (an agent).Thus the deal requires an accurate act of social inference.
Yet, we know little about such inferences. If principals accurately infer how to motivate agents and offer them an attractive deal, then organizations may successfully align the goals of their members; if not, then organizations may fail to meet their fundamental challenge. The social psychology of agency raises an interesting but unexplored question for organizational scholars: How accurately do principals infer the motivations of agents?Some writers have suggested, either explicitly or implicitly, that principals have problems inferring how agents are motivated. For example, Douglas McGregor (1960) explicitly acknowledged problems when he bemoaned the commonness of Theory X managers (who believed that employees dislike work, wish to avoid responsibility, and desire security above all) and the scarcity of Theory Y managers (who believed that employees like work, wish to develop their skills, and desire to participate in tasks that advance worthy organizational goals).Other writers implicitly acknowledged problems.
For example, if managers accurately inferred how employees are motivated, writers would not need to remind them to communicate the importance and relevance of the organization's mission (Hall, 1973; Switchback & Smith, 1993), to provide employees with feedback on their performance (McGregor, 1960), or to use techniques to make Jobs more interesting or meaningful (Hickman & Lolled, 1980; Deeming, 1982).Readers of the management literature might easily conclude that these incorrect inferences are unique to managers, perhaps because they have been socialized inappropriately by their organizations. In contrast, I propose that these incorrect inferences arise because people in general have the wrong theories of motivation. People have intuitive, lay theories about many things (see, e.
G. , Anisette & Ross, 1980; Burnham, 1988), and they also have lay theories about how others are motivated (Miller & Ratter, 1998).When people become managers or principals in organizations, they can act on their lay theories. If lay theories of motivation are incorrect, then people who act as principals may offer ineffective deals to agents.
Thus, lay theories may produce friction when organizations try to meet their fundamental challenge. In this paper, I explore the social psychology of agency relationships by focusing on a particular error in lay theories of motivation, an extrinsic incentives bias.Extrinsic factors are "outside a thing, outward or external"; intrinsic actors are "inward," "belonging to or lying within a given part" (Webster New Universal Unabridged Dictionary, 1994). At a deep level all motivations depend on an interaction between external and internal factors?an extrinsic factor like pay will affect behavior only if individuals have some internal desire for it; an intrinsic factor like "doing something worthwhile" will affect behavior only if there is something in the external environment that individuals consider worthwhile.In the paper, I try to avoid some of these complex issues by defining intrinsic and extrinsic motivation empirically based on the responses of independent sets of observers.
In general, observers in my studies classified motives as extrinsic when they involved an aspect of a situation that would be easily verified by independent Judges (e. G. , pay, benefits, job security) and as intrinsic when they involved an internal state that independent judges might find hard to verify (e. G. An internal change in knowledge like "learning new things" or an internal state of satisfaction like "feeling good about oneself"). However, because motivation depends on an interaction between internal and external factors, even this empirical definition will not completely resolve the issue, and I will revisit this issue at various points in the paper.
Within the management literature, different schools of management have disagreed about whether workers are more motivated by extrinsic or intrinsic factors.Frederick Tailor's scientific management was infamous, in part, because it argued that the deal between organizations and their members should emphasize extrinsic factors: "scientific" managers offered workers better wages and greater security in exchange for working in the most efficient way. Although this efficient work was often menial and intrinsically uninteresting, scientific managers assumed that workers would be willing to make this sacrifice because they were primarily interested in stable, high- paying Jobs: "what workers want most from their employers beyond anything else is high wages" (F.W.
Taylor, 1911). In objecting to scientific management, McGregor and other members of the human relations school of management plausibly argued that scientific managers suffered from a kind obstetricians incentives bias: they overestimated how much employees care about extrinsic task features (like pay or Job security), and they underestimated how much employees were motivated by intrinsic features (like having a meaningful task). 2 For more on this point, see discussion of Study 1 . These lay theories might contain a self-serving bias as well as an extrinsic incentives bias.
The empirical studies in this paper show that the extrinsic incentives bias can be documented even when self-serving biases are controlled. At this point note that, while self-serving biases might explain why we think others are more motivated by money, they seem unlikely to explain 28 CHIP HEATH McGregor and other writers in the human relations school identified the extrinsic incentives bias, but they were less clear about its cause. I propose that this bias is not limited to scientific managers, or even managers in general; it is a property of people's lay theories.In this view, the extrinsic incentives bias is not an occupational hazard; it is a psychological one.
Evidence suggests that an extrinsic incentives bias can be documented in people other than scientific managers. Consider a survey of 486 prospective lawyers, who were questioned by Kaplan Educational Centers during heir preparation for the Law School Admissions Test ("Motives of Prospective Lawyers," 1995). They were asked to describe their own motives for pursuing a legal career and to speculate about the motives of their peers.Sixty-four percent said that they were pursuing a legal career because it was intellectually appealing or because they had always been interested in the law, but only 12% thought so about theorizers.
Instead, 62% speculated that their peers were pursuing a legal career because of financial rewards. Thus, their lay theories stressed their peers' response to money and denied their peers' intellectual interest in the law. In both respects, the lay theories of the prospective lawyers resemble the ideology of scientific managers.Similar evidence for the extrinsic incentives bias can be documented in the population at large. Results from the General Social Survey (ASS) confirm a similar pattern in a randomly sampled group of U.
S. Adults (General Social Survey, 1998). For over 25 years, the ASS has asked a sample of adults to rank the importance of five different aspects of their Jobs: pay, security, free time, chances for advancement, and "important work" that "gives a feeling of accomplishment. " Inevitably, "important ark" is, on average, ranked first (and by over 50% of the individual respondents).Pay typically ranks third. Yet, in the late sass, when the ASS asked respondents about the role of extrinsic incentives for others, people generally believed that pay was quite important.
Of this sample, 73% thought that "large differences in pay' were necessary "in order to get people to work hard," and 67% agreed that "people would not want to take extra responsibility at work unless they were paid extra for it. "3 Combined, these observations suggest that an extrinsic incentives bias may play a role in lay theories of motivation.If so, then McGregor may have erred in describing an extrinsic incentives bias as a characteristic feathery X managers. In their lay theories, members of the general population assess others' motives using Theory X, while they assess their own motives using Theory Y. An extrinsic incentives bias, if it exists, might intrigue organizational scholars because it illustrates a lay theory that might lead principals to craft ineffective deals with agents (and in turn hinder an organization from organizing).
The full pattern of the extrinsic incentives bias exhibited by the scientific managers.Let us start by considering the disposition bias, which focuses primarily on the observers' side of the actor- observer effect. In the traditional attribution experiment, participants observed the behavior of an actor, and then they were asked to explain the actor's behavior (I. E. , attribute it to various causes). In such experiments, observers often overemphasized the actor's intrinsic disposition, a tendency that has been labeled the correspondence bias Cones & Harris, 1967; Gilbert & Malone, 1995) or, more transparently, the disposition bias (Ross & Anisette, 1991).
In the classic ministration of this bias, observers attributed pro-Castro attitudes to a student who wrote a pro-Castro essay, even when they knew the student was assigned the essay topic Cones & Harris, 1967). More recently, observers witnessed another student give a profile or precipice speech on abortion; afterward, they assumed that the speaker held an attitude consistent with his or her speech even though the speech was based on predefined arguments and the observers themselves assigned the topic (Gilbert & Jones, 1986). Thus, observers frequently attribute behavior to dispositions instead of acknowledging the power of situations.This mistake has been documented by investigators in so many situations that it has been called thefundamentalattribution error (Ross, 1977; Ross & Anisette, 1991; Gilbert & Malone, 1995).
In contrast to observers, actors denotation a disposition bias when they explain their own behavior. For example, actors in the experiments above would not explain their behavior in terms of intrinsic attitudes or dispositions (e. G. , Communist sympathies); instead they would attribute their behavior to extern- 30 CHIP HEATH sic, situational factors ("l wrote the pro-Castro essay because it was important for the experiment").Thus, actors tend to emphasize extrinsic attributions for their behavior, while observers emphasize intrinsic ones. Combined, this pattern of attributions has been labeled as the actor-observer effect.
Although researchers have documented some exceptions (see Fiske & Taylor, 1991, up. 72-75), a reader who digested the large literature on attribution could easily conclude that the actor-observer effect is a social constant. However, the lawyer survey and the General Social Survey hint that this empirical result may reverse when people consider agency relationships and explain the workplace or career choices of others.In the traditional attribution study, people say that their own motivations are extrinsic but others' are intrinsic. The lawyers claimed the opposite.
As actors, the prospective lawyers claimed they were motivated by two intrinsic factors: the appeal of a legal career and their longer interest in the law. As observers, they claimed that their peers were motivated by an extrinsic factor: financial rewards. This extrinsic incentives bias differs from the traditional actor-observer effect; thus it may alter our understanding of the attribution process.I propose that the traditional actor-observer effect will frequently reverse when people explain behavior in agency relationships. To understand this proposal, we must first consider why the traditional theory predicts the actor- observer effect in the traditional attribution study and then consider why it might predict an extrinsic incentives bias when observers and actors explain behavior in agency relationships. In explaining the actor-observer effect, the traditional theory has argued that observers and actors differ on two dimensions: perception and information.
First, observers and actors perceive different features of the world.Observers "perceive" the actor (the actor's behavior "is figural or dynamic against a more pallid and dull situational background"; Fiske & Taylor, 1991, p. 73). Actors "perceive" the situation (they notice the situational factors that influence their behavior more than they notice the behavior itself). In response, both observers and actors attribute behavior to whatever is perceptually salient (Storms, 1973).
Second, observers and actors also differ in information. Observers lack information about the actor's past behavior; thus they may infer that actors "always or usually act in this way' (Fiske & Taylor, 1991, p. 3).On the other hand, actors may have information that they have reacted differently in the past, so they resist explaining their behavior as a product of their intrinsic disposition Cones & Anisette, 1972).
While perceptual and informational differences produce an actor-observer effect in the traditional attribution study, the same differences may produce an extrinsic incentives bias in agency relationships. Agency relationships differ from the traditional attribution study in three key features: First, agency relationships involve an explicit deal between an organization and employees, and the deal involves alienate incentives like money.Second, agency relationships are long-term, unfolding over months or years, which means that actors have a chance to habituate to the incentives that are involved. Below, I discuss how these two factors may alter theperceptualdifferences between actors and Defrosters. Third, in agency situations, actors typically make a conscious choices about their Jobs and careers that requires them to consider their own longer preferences and rank their importance.
Below, I discuss how this factor can alter theinformationalasymmetries between actors and observers.
