An economic analysis of the Motivation Crowding-out Theory

Table of Contents

Abstract

Table of Contents

1. Introduction

2. The Motivation Crowding-Out Theory
2.1. Development of the Theory
2.2. Conditions of the Occurrence of Crowding-Out Effects
2.3. Frey’s Theoretical Argumentation
2.4. Implications for Economics Given by Bruno S. Frey

3. Analysis of the Conditions of Crowding-Out
3.1. An Economic Look at Intrinsic Motivation
3.2. An Economic Look at the Expectation of Rewards
3.3. An Economic Look at Performance-Contingency
3.4. An Economic Look at the Perception of Rewards

4. Analysis of the Motivation Crowding-Out Effect
4.1. Economic Fundamentals
4.2. The Crowding-Out Effect in a Broader Sense
4.3. Incentive Structures Designed to Avoid Crowding-Out
4.4. The Old Debate Goes and Will Go On

5. Concluding Remarks

6. Bibliography

Abstract

This paper is concerned with an economic analysis of Frey’s Motivation Crowding-Out Theory, which states that monetary rewards crowd-out intrinsic motivation. Crowding-out effects, conditions as well as implications for economics will be examined in detail by analyzing the highly controversial psychological and economic, mainly human resource, literature. Throughout the paper, the question whether crowding-out effects are of relevance to classical economic theory will be looked at from different point of views. Finally, a conclusion suggests that at this state of information an integration of the concept of crowding-out is not recommendable.

1. Introduction

Monetary rewards crowd-out intrinsic motivation, which is the cause why individuals engage in an activity for its own sake. Undermined intrinsic motivation results thereupon in a decreased work effort level, which in turn leads to higher costs for employers who have to bear the burden of the decreased performance level. Consequently, compensating employees for their work effort is not beneficial for employers.

Obviously the above stated information is not in line with traditional economic theory. The given facts are moreover the crucial points of Bruno S. Frey’s “Motivation Crowding Theory: A Survey of Empirical Evidence”.^[1] Frey is professor of economics at the University of Zurich and he is best known for his critiques of the Homo economicus. He was one of the first economists who argued that “the crowding-out effect is one of the most important anomalies in economics”.^[2]

This paper is concerned with an economic analysis of Frey’s Motivation Crowding Theory (MCT). ‘Possible’ crowding-out effects and their relevance for economics are the focus of research throughout the present paper. This contradiction is dealt with by analyzing recent economic literature as well as the more traditional school of human resource management. Caused by the inconsistent literature, different questions arose during the examination of existing research argumentation: Can the existing empirical evidence be applied to economics? Does Frey’s approach demand the integration of psychological concepts into economics, or does it make it even mandatory? In order to draw strong conclusions from the given evidence in the end, the following procedure was applied: First and foremost, the argumentation and implications on the psychological level, which back up Frey’s findings, are accepted unquestioningly within this paper. Also crowding-in effect will not be dealt with in detail, as it will be exposed in the second chapter.

The upcoming chapter of this paper precisely summarizes Frey’s MCT. Starting from a short introduction of the development of the theory, the four most significant conditions which have to be fulfilled in order to prove the occurrence of a crowding-out effect will be explicitly clarified. Adopted from Frey’s approach, his argumentation method employed will be shortly reproduced to eliminate indistinct fundamental ideas. As a last issue, Frey’s implications for economics will be summarized. This second chapter will function as a rich foundation for the subsequent economic analysis.

The chapter afterwards analyzes the economic applicability of the conditions of Frey’s MCT. The conditions that have to be satisfied will be analyzed in the following manner. Firstly, it will be shown that employees are not always intrinsically motivated. Secondly, it is demonstrated that employees do indeed expect monetary rewards in exchange for their labor. Thirdly, pay-for-performance schemes, which are contingent on the outcome of performance, are examined. Fourthly, conclusions will be drawn about the perception, either as controlling or as supportive, which depends mainly on the way rewards are carried forward to employees.

The economic analysis in the last chapter will be based on the basic economic assumption, that performance is positively related to effort and that „effort is unpleasant and money is good“.^[3] As long as the utility, which is mainly acquired from being compensated for one’s effort, exceeds the disutility an individual experiences from working, the respective individual will be motivated to engage in such a trade-off of performance and pay. The highly debated crowding-out effect will be analyzed and clarified in greater depth, by opposing psychological with economic literature.

As a final result, the suggestion that, based on the given classical assumptions, the crowding-out effect does not hold for economic theory will be given. MCT is a real phenomenon, however under the given conditions it cannot be applied as easily as Frey proposed it. Until complete clarity is not given an integration of psychological concepts into economics is not advisable.

2. The Motivation Crowding-Out Theory

In his MCT, Frey strives for the goal to integrate psychological concepts into economics. He strongly argues that rewards crowd-out intrinsic motivation. Such rewards can be positive monetary ones or organizational regulations that are accompanied by negative sanctions. A crowding-out effect is most likely to occur if monetary rewards are involved. Based on this finding and on the immense meaning of money in contemporary life, the present paper focuses only on the effects of monetary rewards on intrinsic motivation. The underlying conditions of this effect, as well as arguments and findings that support Frey’s MCT will be described in the upcoming chapter.

2.1. Development of the Theory

Frey is one of the first researchers that specifically addressed the issue of motivation effects, in particular intrinsic motivation effects, against the background of economics. His approach emanates from research of cognitive social psychologists. Deci, who is a professor of psychology at the University of Rochester, set already the basic elements in 1971, when he argued that money has a loosing effect on intrinsic motivation for the respective activity.^[4] He based his arguments on research and experiments in which monetary rewards were paid to college students for solving simple tasks. In Deci and Ryan’s Cognitive Evaluation Theory, the two researchers specified that the undermining effect of monetary rewards depends on the interpretation of the reward-receiver and the main focus are explicit conditions that undermine versus enhance intrinsic motivation.^[5] Until today, numerous experiments have extended and specified Deci’s findings. Lepper and Greene, for example, named the initial and still controversial negative effect of rewards on intrinsic motivation the “Hidden Costs of Reward”,^[6] and finally Frey termed it the “Motivation Crowding Effect” in 1997.^[7]

Considering the just mentioned highly debated and controversial interplay of intrinsic motivation and monetary rewards, the following two effects are possible outcomes:

1. Monetary rewards crowd-in (strengthen) intrinsic motivation.
2. Monetary rewards crowd-out (undermine) intrinsic motivation.^[8]

It may be noted here, that effects which strengthen intrinsic motivation will not be discussed in detail within this paper; nevertheless they will be mentioned throughout the explanation of theory. Crowding-in effects are in line with economic theory and moreover if the existence is proved they are less troublesome, in fact they are desirable. What is more interesting, for economists, is the reputedly proved crowding-out effect, which is conflictive with economic theory. This problem will be addressed by following up on Kreps’ posing of the question: “Without extrinsic incentives, why would a worker expend any effort?”^[9] Kreps thereby refers to the reality of designing efficient compensation packages, which mainly consist of monetary rewards to sort, attract and retain employees. In order to finally answer the central questions in the concluding section of this paper, within the next chapter, the conditions of the occurrence of crowding-out effects will be accurately listed, based on these conditions a more detailed treatment in terms of economic applicability will follow in chapter 3. In this way, the economic relevance of the crucial characteristics of rewards with a potential crowding-out effect will be determined in the 4th chapter, followed by concluding remarks which readdress the questions posed above.

2.2. Conditions of the Occurrence of Crowding-Out Effects

The MCT suggests that monetary rewards undermine or strengthen intrinsic motivation. Frey brings forward the argument that crowding-out effects create a negative relationship between extrinsic and intrinsic motivation. Money might lead agents to reevaluate an activity which was intrinsically motivating before and becomes one which is – after having introduced a reward - primarily motivating by the anticipation of money.^[10]

Certain conditions have to be met to finally provide evidence for either of the two effects. Crowding-out as well as crowding-in of intrinsic motivation requires the most important condition for Frey’s theory, namely that there was intrinsic motivation in the first place.^[11] If intrinsic motivation is not present, economic theory surely holds since it is the same situation as it would be treated exogenously. In this case, performance increases with higher monetary rewards according to the relative price effect. However, intrinsic motivation is taken for granted throughout Frey’s experiments.

Yet, there are three more conditions, which have to be fulfilled in order to satisfactorily confirm a crowding-out effect. “Rewards do not crowd-out intrinsic motivation, when they are unexpected or not contingent on task behavior”.^[12] Frey again derives those two conditions from psychological research investigated by Deci, Koestner and Ryan.^[13]

Secondly, unexpected rewards do not have to be considered. Verbal rewards, for instance, are in almost all cases unexpected during task performance and do not significantly effect intrinsic motivation. Moreover they tend to enhance people’s feelings of competence, which strengthens intrinsic motivation.^[14] Therefore as another criterion, intrinsic motivation can only be crowded-out if the employee did expect the reward in advance.

Turning back to the characteristics of rewards that crowd-out intrinsic motivation, thirdly performance non-contingent rewards do not fall into the category of relevant required conditions. Such typical input-based rewards are administered in a way that they are perceived as supportive while they are based upon the level of effort or time and not upon the level of output. On the other hand, performance-contingency emphasizes that, by paying based on the output, the activity is instrumental for attaining rewards, also it is indicated that individuals have done well enough to receive the reward. Paying employees by the output they have produced strongly controls them, since the output will be measured and furthermore these employees are directed from the outside. Alfie Kohn observed the same effect: “The bottom line is that any approach that offers a reward for better performance is destined to be ineffective”.^[15]

Based on psychological findings, the appearance of crowding effects also depends on the way the reward is perceived by the receiving individual. Rewards have two possible aspects, “[…] a controlling aspect – the aspect that controls or regiments behavior – and an informational aspect – the aspect that conveys positive or negative information about a person’s competence and self-determination”.^[16] If rewards are observed as supportive, the informational aspect is leading. This can be due to a higher need for information or to the way the reward is assigned and administered, thus the feeling of internal control of the reward receiver is strengthened and the feeling of self-determination is changed. The intervention does support perceived autonomy or competence, which enhances intrinsic motivation.^[17] On the other hand, the feeling of being directed from the outside can arise; this is the case if the reward has controlling character. Thereby the experienced external control of the recipient is strengthened and furthermore intrinsic motivation will be crowded-out, because the locus of control shifts from the inside to the outside.^[18] Summarizing, the recipient’s intrinsic motivation will only be undermined if his “perception is that the controlling outweighs the informing effect”.^[19]

Recapitulating the interplay between intrinsic and extrinsic motivation, once more mention should be made of the conditions of a possible crowding-out effect. Intrinsic motivation can only be undermined, if intrinsic motivation is existent in the first place (see chapter 3.1.), the reward was expected (see chapter 3.2.) and is contingent on performance (see chapter 3.3.) and as a last criterion the reward has to be perceived as controlling by the receiver (see chapter 3.4.). Based on the fact that the conditions are complied with Frey corroborates his theory by giving powerful theoretical arguments that challenge, for instance, the law of demand. The following subchapter reveals Frey’s approach toward the phenomenon of crowding-out of intrinsic motivation.

2.3. Frey’s Theoretical Argumentation

Based on the assumption that all four conditions, which are mentioned above, are met Frey and Osterloh attributed the crowding-out effect to five underlying sub-effects that complement each other, these can be shortly summarized as: impaired self-determination, reciprocity, fairness, reactance and spillover effect. Since these effects mainly rely on psychological concepts and do not help to answer the guiding question, they are only mentioned within this paper without giving a comprehensive explanation.^[20]

Keeping in mind the four conditions of the occurrence of the crowding-out effect and the five mentioned sub-effects, it is now interesting to progress to Frey’s economic approach demonstrating the effect of monetary rewards on performance. Frey argues against the law of demand, which states that if the price of a good is raised (costs of an activity), demand for that good decreases (or the respective activity is reduced). People adjust to the new situation of given constraints. In personnel economics the same holds, monetary rewards and commands impose a relative price effect on the agent, also known as the “disciplining effect”.^[21] This is the point where Frey draws opposite conclusions. According to him, if the costs of an unwanted activity are raised (higher monetary reward levels impose higher costs on shirking) then demand does not decrease as expected, it increases due to a change in the type of motivation, from being intrinsically motivated to being extrinsically motivated. Whether this effect is justifiable depends, according to Frey, again on the outcome whether external intervention strengthens or undermines intrinsic motivation, resulting in the question whether psychological concepts should be integrated into economic theory. In the first case of crowding-in, the effect of disciplining the agent is further strengthened by the crowding-in of intrinsic motivation. Crowding-out, on the other hand, negatively affects the agent’s marginal benefit from performing, resulting in a lower performance level of the agent. Thus, the disciplining effect is not applicable in this case.^[22] Frey argues, that in general “both the relative price effect and the crowding-out effect are active, so that external intervention has two opposite effects on the agent’s performance”.^[23] The principal’s intervention might have a negative effect on an agent’s performance. This is the case if the well-known upward-sloping supply curve is shifted to the left and as a consequence, the work effort level of the agent decreases, although monetary rewards are increased.

The relative size of the disciplining effect and the crowding-out effect is the crucial factor whether it is beneficial for the employer to intervene. She or he^[24] has to identify those rewards that on the one hand do not decrease intrinsic motivation, moreover those which provide an increase in utility for the employee, and on the other hand cause low costs for the employer. In this case the costs associated with a lower work effort level due to a possible crowding-out of intrinsic motivation have to be kept down.^[25] The difficulty of cost-utility analysis of compensation packages will be readdressed in chapter 4.2. Resuming and concluding the analysis of Frey’s crowding-out effect, implications for the principal-agent problem and to a greater extent for economics in general, will be given in the next chapter.

2.4. Implications for Economics Given by Bruno S. Frey

Having examined development, crowding-out conditions and argumentation of Frey’s approach, importance should now be attached to the resulting consequences for economics. Crowding effects in general are potentially relevant; however crowding-out effects possess a more significant function in economic theory, since crowding-in effects are not regarded as controversial issues due to the strengthen of the relative price effect. An ‘affected’ area of particular interest for motivation crowding-out research is the field of personnel economics. Accepting crowding-out as proven (opposed to economic theory), higher compensation would than implicate a lower level of effort, involving principal-agent theory and all its implications are not reliable any longer. Therefore Frey questions all instruments and incentives that rely on the price effect. In his opinion, such views are only supported by theoretical economic literature, thus he criticizes the generalization of conclusions resulting from such empirical studies. Generalizations are too hastily made in simple task environments where intrinsic motivation is irrelevant and therefore can not be crowded-out. Frey is of the opinion that it is necessary to reconsider the limits of the firm with regard to the possible limits resulting from purely relying on monetary incentives. Not surprisingly, as a next step, pay-for-performance, which he describes ironically as “the uncontested instrument for eliciting higher performance from employees”,^[26] is criticized, as the power of payment-based measures is overrated. In his MCT, he refers to several experimental researches to warrant his implications. To mention one, Fehr and Gächter tested the hypothesis that incentive contracts crowd out voluntary cooperation.^[27] Agents determine their effort choice by maximizing their utility, which means to find the best possible combination of effort and income that keeps the costs of effort low and that maximizes their income at the same time. On the other hand, one speaks of voluntary cooperation if the agent spares no effort to yield a higher output, although his marginal benefit of effort is not affected. This can be attributed to irrationality or the agent may just have an intrinsic preference for the activity or the agent may feel an obligation to exhibit loyalty because the principal has been treating him well on a personal level.^[28] Fehr and Gächter’s experiments show that “there are incentives that have counter-productive effects on fairness-driven voluntary cooperation and there are incentives which enhance such voluntary cooperation”.^[29] This raises special problems, but it should not be the question whether to use or not use monetary rewards in general, it is rather the task to identify fairness compatible monetary incentives that do not crowd out intrinsic motivation and in this case voluntary cooperation, as Fehr and Gächter conclude. Their findings mainly confirm Frey’s crowding hypothesis. The experiment observed a trust-treatment and an incentive-treatment, which was based on fines. Due to a higher voluntary cooperation of the agent the total surplus is the highest in the trust-treatment compared to the incentive-treatment. Their findings also indicate that ‘incentive-compatible incentive treatments’ (as they are called throughout their experiment) are much more profitable for principals than non-incentive-compatible ones, i.e. where the no-shirking condition is violated. The crowding hypothesis is confirmed for incentive-compatible contracts, because in this case the average excess effort is virtually never positive, i.e. there is nearly no voluntary cooperation.^[30] Consequently, incentives should be designed in such a way that they do not function counterproductively to voluntary cooperation and therefore do not crowd-out intrinsic motivation.

Subsequent to the examination of the empirical evidence, Frey concludes that his “survey shows that […] skepticism is unwarranted and that strong empirical evidence indeed exists for crowding-out and crowding-in”.^[31] He also adds, that crowding effects “do not always prevail over the traditional relative price effect”.^[32] Although Frey remains of the conviction that his theory is an essential contribution to modern economics, as a final remark he points out that the identified conditions have to be more clarified and specified in future research. Following up Frey’s remarks the two upcoming chapters mainly addresses the question whether the experiment design and Frey’s assumptions are generally applicable to economics, in particular personnel economics. Since Frey’s MCT is still mainly based on psychological concepts, the present economic analysis touches the conflicting interests and arguments of economics and psychology. It remains subject of future research to give more far-reaching, detailed and profound economic explanations to this topic.

3. Analysis of the Conditions of Crowding-Out

The underlying fundamentals of the motivation crowding-out theory demand further explanations concerning its relevance for economics. By the use of economic terminology and principles the crowding-out effect will be analyzed in detail by splitting the outlined conditions into their component parts, followed by an economic analysis of those parts in order to finally draw comprehensive and critical conclusions about Frey’s economic approach of the crowding-out effect. Recalling these conditions, they will be analyzed in the following manner: Intrinsic motivation has to be existent in the first place, the monetary reward has to be expected, it has to be contingent on performance and the employee has to perceive it as controlling.

3.1. An Economic Look at Intrinsic Motivation

Creating Motivation seems to become one of the most significant tasks of management. Creating Motivation – this involves giving reasons to your employees to increase performance and efficiency and to invest higher efforts in the long run. One goal of most incentive contracts is to generate the ‘right’ motivation. But what kind of motivation is the desired one? How can we influence people in order to make the company’s goal to their own individual goal? What can we actually understand under motivation, and in particular intrinsic and extrinsic motivation? This section of the paper is going to clarify the terminology about intrinsic and extrinsic motivation in an economic context in order to answer the question whether employees are intrinsically or extrinsically motivated.

Due to the complex nature of motivation, it does not exist such as a single rightful definition of motivation. A simple definition given by the social-psychologists Deci and Ryan can certainly be applied to economics: “Someone who is energized or activated toward an end is considered motivated”.^[33] Motivation can be traced back to an unsatisfied need. As soon as the need is satisfied, it cannot be the driving force behind human behavior anymore.^[34] There are many driving forces which can, in fact, be the reason for certain actions, e.g. power, effort, safety, order or comfort. Since these driving forces are rarely perfectly aligned with the interests of the company, motivation questions may arise. Based on Bruno Frey’s research, motivation is not an end in itself; moreover it ought to serve the objectives of the company.^[35] In the organizational context, Robbins defines motivation as follows: “Motivation is the willingness to exert high levels of effort toward organizational goals, conditioned by the effort’s ability to satisfy some individual need”.^[36] Thus, effort should be consistent with the goals of the organization. This can often be easily accomplished by introducing monetary incentives. Most employees are very motivated on the first working-day, however the initial enjoyment and pleasure declines in the course of time. Giving your employees reasons for high effort levels can be accomplished by efficient incentive contracts.^[37] Such contracts have to take into account the motivation costs, which are all costs that assure the best-possible performance and execution of duties and responsibilities assigned to the agent.^[38] However, agents only agree on the contract if the personal advantage is obvious, they have to be convinced that from their point of view it is only beneficial to perform well or at an even higher level.^[39]

Extrinsic motivation “serves to satisfy indirect or instrumental needs”.^[40] Money can nearly almost be regarded as a means to an end. Work is just an instrument which yields satisfaction of one’s actual needs by the means of compensation. In order to afford a new car or a vacation one has to work, while work is not the end. By receiving pay for one’s performance one can pay the long looked-for vacation, which is the real need satisfaction at the end.

In contrast, a person is intrinsically motivated if the activity is “an end in itself”.^[41] Intrinsic motivation can further occur in the form of job satisfaction, achievement of personal goals or one just wants to comply with standards for their own sake.^[42] An important pre-condition for being intrinsically motivated is that the respective person has chosen the values himself and accepts them to be relevant for his own behavior. In economic terms, an agent performs an activity because it is subject of his utility function. Decisive for the existence is the task itself and the preference structure of the employee. Intrinsically motivated workers feel that “their work behavior is self-fulfilling”.^[43] Employees who enjoy what they are doing at work have a preference for exactly those tasks. Such employees either demand less pay or they perform better at the same pay level compared with extrinsically motivated employees, or they produce higher quality or generate higher levels of output.^[44] Having briefly summarized what intrinsic motivation means in the economic sense, its significance and the question whether it bears a meaning in economic theory will now be analyzed.

To readdress the guiding question whether employees are intrinsically motivated or not, it has to be pointed out that according to most economists managers normally cannot count on employing intrinsically motivated workers. It is assumed that individuals can only be convinced to act in favor of the organization, by remuneration and punishment, profit and loss, temptation and pressure, shortly: through incentives and regulations.^[45] Within this context, it is beyond question that intrinsically motivated employees are more beneficial to companies, since they cooperate voluntarily. It is moreover the task to determine whether employees are intrinsically motivated at all.

According to economic and psychological definition, intrinsic motivation refers to an activity that is interesting and is performed for its own sake. However following economic literature, the interestingness of the required performance is not ranked first. In fact, the compensation package is an almost all-dominant criterion for one’s choice of effort. Money is in almost all employer-employee relationships a decisive measure which allows the fulfillment of primary needs, e.g. food, water or security of a home. Expressing it simplified, intrinsically motivated employees need to eat, too. If monetary compensation is not assured, than it is also not typical for employees to engage in economic interactions. However, if it is assured than it still depends on the job design. Assuming that the decisive factor of income is agreed on, the intrinsic motivation of assembly line workers might be lower than that of a commercial artist. Job design can be influenced by many factors. Job enrichment programs, for example, redesign employment characteristics to add variety to monotonous jobs. Consequently, these tasks will also gain in intrinsic value by making them more appealing to employees. Job satisfaction and production can be increased by providing the missing variability and challenge.^[46] However, employees who work only because of the joy derived from their job will certainly not be motivated by more money, for instance employees of a non-profit organization. In this case other, than monetary, motivation techniques should be used in order to enhance the motivation of those employees.

[...]

---

^[1] See Frey/Jegen (2000), p. 1. The working paper was published by Frey and Jegen, but it is known as Frey’s Motivation Crowding Theory. Therefore it will be referred to Frey’s Theory within this paper, too.

^[2] Frey/Jegen (2000), p. 3.

^[3] Gneezy/Rustichini (2000), p. 791.

^[4] See Deci (1971), p. 114.

^[5] See loc. cit., p. 105.

^[6] Greene/Lepper (1978), p. 109.

^[7] Frey’s Motivation Crowding Theory was first published in 1997, but this paper is going to analyze the revised version of 2000.

^[8] See Frey/Jegen (2000), p. 6.

^[9] Kreps (1997), p. 360.

^[10] See Muchinsky/Turnage (1976), p. 482.

^[11] See Frost/Osterloh/Weibel (2002), p. 16.

^[12] Frey/Jegen (2000), p. 12.

^[13] See Deci/Koestner/Ryan (1999), p. 628.

^[14] See Deci/Ryan (2000), p. 22.

^[15] Kohn (1993), p. 119.

^[16] Deci/Porac (1978), p. 162.

^[17] See Deci/Ryan (2000), p. 17; and Falk/Fehr (2002), p. 37.

^[18] See Jost (2000), p. 506; and Frey/Jegen (2000), p. 9.

^[19] Frey/Osterloh (1997), p. 319.

^[20] See Frey/Osterloh (1997), p. 313-315.

^[21] Frey/Jegen (2000), p. 6.

^[22] See loc. cit.

^[23] Frey/Jegen (2000), p. 7.

^[24] From now on only ‘he’ will be used, which stands for she or he.

^[25] See Lazear/Wolff (2001), p. 225.

^[26] Frey/Jegen (2000), p. 10.

^[27] See loc. cit., p. 9, 16.

^[28] See Fehr/Gächter (2001), p. 3.

^[29] See loc. cit.

^[30] See Fehr/Gächter (2001), p. 23-24.

^[31] Frey/Jegen (2000), p. 23.

^[32] See loc. cit.

^[33] Deci/Ryan (2000), p. 54.

^[34] See Brandt/Reece (1994), p. 117.

^[35] See Frey/Osterloh (1997), p. 315; and Frey (2001), p. 2.

^[36] Robbins (1996), p. 212.

^[37] See Lazear/Wolff (2001), p. 49.

^[38] See Jost (2000), p. 519.

^[39] See Wolff (1999a), p. 44.

^[40] Frey/Osterloh (2002), p. 8.

^[41] Falk/Fehr (2002), p. 35.

^[42] See Frey (2001), p. 25; and Deci/Ryan (2000), p. 56.

^[43] Aldag/Brief (1977), p. 498.

^[44] See Lazear/Wolff (2001), p. 13.

^[45] See Scherhorn (1999), p. 123.

^[46] See Milgrom/Roberts (1992), p. 411.