Lade Inhalt...

The Impact of Financial Incentives on Individual Performance: An Experimental Approach

Diplomarbeit 2010 80 Seiten

Führung und Personal - Mitarbeitermotivation



Declaration of Authorship


List of Tables

List of Figures

1 Introduction
1.1 Scope of Research
1.2 Outline of the Thesis

2 Theoretical Background
2.1 Psychological Background
2.2 Economic Theory

3 Research Findings
3.1 Empirical Findings
3.2 Experimental Findings
3.3 Critical Acclaim and further Questions

4 Proposals for the Structure of an Experiment
4.1 Objective
4.2 Design
4.3 Organization
4.4 Subsequent to the Realization
4.5 Potentially occurring Challenges

5 Conclusions
5.1 Summary of the Results
5.2 Further Research Suggestions

Appendix A

Appendix B


Declaration of Authorship

I hereby confirm that I have authored this diploma thesis independently and without use of others than the indicated resources. All passages, which are literally or in general matter taken out of publications or other resources, are marked as such.

Steffen Hetzel

Mannheim, February 19, 2010


The thesis on hand is dealing with the impact of financial incentives on individual performance. For this, the perception of an experimental approach has been chosen. The target of the thesis is the development of the blueprint of an experiment to provide further research input on the effectiveness of financial incentives.

To do so, the theoretical background for studying this problem is introduced by investigating the psychological and economical approaches to analyze the topic. Additionally, empirical and experimental studies dealing with this issue are presented.

Based on those findings, the structure of an experiment to be carried out at university with students is developed and objectives, design and supplementary requirements for conducting this are discussed. Subsequent, suggestions for the analysis, reporting and possibly occurring challenges throughout the process of implementation are illustrated.

The design of the experiment is giving a verification of before detected findings of a non-linear correlation between incentives and performance. In contrary to standard economic models, the relation is not predicted to be monotonic, but S-shaped. For this perspective, not only performance on varying incentive levels is analyzed, but also performance if payments are absent. Furthermore, the influence of publishing the course of incentive levels in the beginning of the experiment, in comparison to a task-to-task announcement is investigated. An evaluation of this relation is undertaken by studying the impact of financial incentives on performance of three observation groups through two different exercises with varying incentive levels during a real-effort experiment.

List of Tables

Table 1: Incentive Structure of the Experiment conducted by Ariely and Heyman

Table 2: Summary of described Experiments

Table 3: Structure and Compensation of Experimental Groups

Table 4: Advantages and Disadvantages of Laboratory Facilities

List of Figures

Figure 1: Conceptual Framework of the Effects of Monetary Incentives on Effort and Task Performance

Figure 2: Structure of Organismic Integration Theory

Figure 3: Illustration of the Crowding-Out Effect of Intrinsic Motivation

Figure 4: Illustration of the Backward Bending labor Supply Curve (own illustration)

Figure 5: Results of Experiments from Ariely and Heyman (2004) (own illustration)

Figure 6: Suggested Relation of Incentive Level and Individual Performance

Figure 7: Example for Counting Numbers Task

Figure 8: Schematic Representation of Sliders

Figure 9: Example for Slider Task

Figure 10: Example for Matrix Task

1 Introduction

Many jobs are today paid via performance-based payment or at least have a component including financial payments based on individual performance. Especially, as ones performance is comparably easy to investigate, financial incentives dependent on visible achievements seem useful and practicable and in consequence are multiply used in today’s society. This means, that numerous employees, especially at a particular level (Managers) or with a substantial part of identifiable success (Sales, Consulting) are rewarded for their effort based on observed measures of performance. The intention behind performance-based compensation is to stimulate individuals to increase their motivation and effort spent on tasks, and hence their output, or in other words, profitability for the company. By using this way of payment, special incentives are placed and a change in ones behavior is intended to be reached for receiving the maximum possible outcome for the company.

On the other hand, the usage and effectiveness of financial incentives is not always seen as a positive instrument at all. The application of those tools and resulting behavior changes are seen as one of the major reasons for the financial crisis beginning in 2007[1]. Wrong incentive setting for the employees and hereby misleading behavior of the staff yielded to specific behavior which lead to a fundamental crash of the world economy and threw the world into a global recession.

After those incidents, the mode of functioning of financial incentives and the most favorable usage of this motivation instruments came forcefully to the spotlight. Especially, comparably high “bonus payments” for bankers were in consequence of billions paid by governments for bank rescue and avoiding a collapse of the global financial system publicly discussed.

1.1 Scope of Research

The existing and fundamentally logical point of view on incentives is that incentive payments lead to increased motivation which resulted in higher performance of an individual. Furthermore, a positive correlation between incentives and performance is supposed to exist. This means, the higher incentives are, consequently the higher the motivation for conducting a task and followed by this the performance on a task can be expected to be.

The need of designing optimal contracts has always been a challenge, not only for companies. Economical research is delivering tools for understanding and minimizing problems arising through incomplete contracts between employer and employee. Previous to the design of optimal contracts, the efficiency of financial incentives needs to be investigated and proven.

Psychological research is laying the focus on analyzing motivation and performance of employees (e.g. Maslow, 1943; Herzberg et al., 1957; Deci and Ryan, 1985) and how the connection between intrinsic and extrinsic motivation can be affected through the application of different rewards.

From an economic perspective, the Principal-Agent-Theory is dealing with the relationship between employee and employer (e.g. Mirrlees, 1976; Grossmann and Hart, 1983; Holmstrom and Milgrom, 1987) and suggests solutions for minimizing the problem of asymmetric information and consequently providing optimal contracts for receiving maximum performance from the employee.

Theoretical approaches on applying small financial incentives in comparison to their absence have been analyzed (e.g. Frey, 1997; Bénabou and Tirole, 2003; Fehr and Falk, 2002) and supplementary, as incentives increase, the effects on individual willingness for achieving additional earnings has been investigated (e.g. Gilbert and Pfouts, 1958; Hanoch, 1965; Kahneman and Tversky, 1979). However, those models are hardly to be reviewed via empirical analysis as data are difficult to survey[2].

An experimental surrounding is predestined for the issue of analyzing and deepening findings on individual behavior in a tightly controlled environment and by this enriching economic research. So far, conducted experiments give useful information on the effectiveness of varying financial incentives (e.g. Gneezy and Rustichini, 2000a; Ariely and Heyman, 2004; Ariely et al., 2009; Pokorny, 2006) but do not come up with an explicit explanation on the width mode of functioning.

The thesis on hand is enlarging the branch of research by suggesting the design of an experiment for further analysis of the way financial incentives work. The approach of experimental research is chosen in consequence of the advantages in using experiments for analyzing people’s behavior and in testing suggested theoretic models under supervised conditions[3].

A special perspective has been laid on lately conducted experimental studies to offer a current insight into the state of research. Additionally, the theoretical backgrounds on incentive effectiveness are introduced, which include psychological research on motivational development into economical optimization constraints.

The scientific relevance of this thesis is justified through the target of enhancing the branch of research of experimental economics by expending the perception on the effectiveness of financial incentives on individual performance. Consequently, it features assistance on promoting the understanding of individual behavior and principal-agent problems and enlarges the findings on analyzing the incentive-performance relation from an experimental perspective. By suggesting the design of an experiment to be conducted, the bases for subsequent analysis are being laid and successive research is relieved.

1.2 Outline of the Thesis

For the purpose of enlarging the research, this thesis is arranged in four parts:

First, an intuition on the theoretical research on the effectiveness of financial incentives is given. For this purpose, Chapter 2 provides an overview of the psychological literature, followed by an introduction of economic theory.

In Chapter 3, performed empirical studies by means of analyzing data sets and findings from realized experiments on financial incentives are described in more detail.

Based on the findings of those recently conducted experiments, a guideline for the body of an experiment for further research is introduced in Chapter 4. Therefore, results of the presented researches are compiled and the possible course of a graph expressing the relationship between financial incentives and individual performance is displayed. Questions that arose through conducted studies are pointed out and hypotheses for further analysis are introduced. Adapted from the hypotheses, the design of an experiment for enlarging the research is displayed. In connection with this, the requirements for the execution of the experiment are introduced and discussed. Suggestions for analyzing collected data within the experiment and reporting raised findings are listed. In the end, possible occurring challenges throughout the procedure of the experiment are mentioned and potential options for tackling them are suggested.

In Chapter 5, conclusions on the introduced analysis are drawn and expected research findings as well as further research areas are pointed out.

2 Theoretical Background

The analysis of individual performance and effort by empirical and experimental research is one important point to achieve results helping to understand the mode of functioning of financial incentives. A second point is the bedding of those results with theoretical backgrounds or models. In psychological research, the connection between experimental and theoretical research is (so far) closer than in economics, where the spotlight more relies on theoretical and empirical research. “Economics for the last forty years and more is strongly theory-based. Acceptable economic theories must be fully developed from preestablished first principles, and must relate specifically to the core micro theory” (Friedman and Sunder, 1994, p.133).

illustration not visible in this excerpt

Figure 1: Conceptual Framework of the Effects of Monetary Incentives on Effort and Task Performance[4]

As drawn in Figure 1, effort is dependent on Motivational Mechanisms and varying specific factors of influence (such as personal skills, task difficulty or environmental variables). Monetary incentives can influence individual motivation, through this impact effort, and consequently task performance. As performance on a task is heavily dependent on individual factors of influence which are difficult to appreciate upfront, theoretical constructs focus on analyzing motivation and effort (see Maslow, 1943; Herzberg et al., 1957; Deci and Ryan, 1985; Frey, 1997; Bénabou and Tirole, 2003). In the following, the terms and definitions “effort” and “performance” are used simultaneously, as this is most practicable for the analysis. Nevertheless, the differentiation as shown in Figure 1 is important to know and should be kept in mind by the reader.

In the first part of this chapter, an introduction into the psychological approach of analyzing individuals’ motivation (and by this one’s performance) is given.

The second part of this chapter lays the focus on the theoretical research economists contributed to the question of how financial incentives influence ones behavior. And continues on how the problem of optimizing individual preferences can theoretically be solved by using specific mechanisms.

2.1 Psychological Background

Based on the close interaction between experimental and theoretical analysis in psychology, a considerable background on this research area has been built up. Numerous theoretical models on the effects of financial incentives on individual performance have been delivered. The main focus of research is relying on analyzing the effects, incentives exercise on intrinsic motivation and in which way this interaction influences individual motivation and performance. One focus of this subchapter relies on the relationship of effort on a task and the development of this concerning individual motivating factors.

In psychology the view on individuals’ performance is biased by two factors. One is being stimulated by a motivating factor from outside (e.g. rewards, positive feedback), namely extrinsic motivation, and the second one is performing a task from an inner (intrinsic) motivation. Financial incentives impact an individuals’ motivation and the performance of the recipient. However, the effects are not determined and can work in a positive as well as a negative direction. In psychological literature, a general contention in many textbooks and journal articles is that rewards can be harmful to an individual's intrinsic motivation and by this a crowding-out of overall motivation can take place (Deci, 1975; Deci and Ryan, 1985; Lepper and Greene, 1978).

As studies (and theoretical constructs) are more focused on the effects on intrinsic motivation, empirical analysis must be reviewed in a different way than comparable economic research. Conducted meta-analysis provided no definite results in the way of functioning of extrinsic motivating factors[5] (compare Rummel and Feinberg, 1988; Wiersma, 1992; Tang and Hall, 1995; Cameron and Pierce, 1994; Deci et al., 1999; Eisenberger and Cameron, 1996). Positive as well as negative impacts have been detected and depending on the analysis, the results differ accordingly. There is little doubt about external factors impacting individuals’ motivation, but the modality is questionable and not definitely obvious.

The Relation of Extrinsic Factors and Individual Effort

As early as 1908 Yerkes and Dodson (1908) analyzed the effects of extrinsic factors on individual effort. For this, the authors investigated the connection between activation and performance throughout an experiment with mice which were rewarded depending upon their behavior on learning discrimination between different grey papers. Modeling the experiment helped to analyze the connection between motivation and performance. The authors found a relationship between arousal and performance which can be described as an inverted U-shaped function with a peak after which the mice had a slower or even decreasing learning curve. Those findings are today known as the so-called “Yerkes-Dodson” law. The point of the optimum of arousal is expected to vary depending on the task, the individual's personality, and the individual's experience with the task. Since arousal is closely connected to motivation and performance, the “Yerkes-Dodson” law implies that increases in motivation beyond an optimal level can, in some situations, produce a supra-optimal level of arousal and hence decrements in performance (compare Ariely et al., 2009, p.452).

Follow-up experiments showed, that increased motivation can limit individuals’ focus of attention on a variety of dimensions, together with the extent of the solution set people consider (e.g. Easterbrook, 1959) and caused by this, results in a negative effect on performance[6] (see Ariely et al., 2009, p.453). By showing that a negative impact on performance was realized through the introduction of monetary incentives for tasks that involved problem-solving, McGraw and McCullers (1979) supported the theory of this mechanism.

One explanation for those effects can be seen in large incentives to occupy the mind and concentration of the job holder with thoughts about her future and if she will get the payment and her regrets if not, are diverting her from the task at hand. Further explanations for negative effects of increased incentives on individual effort might be, that increased incentives can cause people, involuntarily, to intentionally think about the task and by this shift control of behavior from “automatic” to “controlled” mental processes although it is well known that controlled processes are less effective for tasks that are highly practiced and automated[7] (see Ariely et al., 2009, p.453).

Motivation Theories

At least three historical theories of motivation have dampened the attention of psychologists in investigating the relations between performance, performance evaluation, and pay (compare Rynes et al., 2005, p.574): Maslow’s hierarchy of needs theory (Maslow, 1943), Herzberg’s motivator-hygiene theory (Herzberg et al., 1957), and Deci and Ryan’s Self Determination Theory (Deci and Ryan, 1985)[8]. A major implication of all three theories is that monetary rewards are not a major determinant of work motivation, except perhaps for employees at low income levels.

In Maslow’s need hierarchy theory (1943) human needs are arranged in a hierarchy of pressure that is biological or instinctive in nature. Consequently, a need that is denied acts as a primary motivator; while a need that is satisfied has less motivational impact.

The base of Maslow’s hierarchy is built upon physiological (food, sleep) and safety (e.g. housing) needs - needs which are most effectively settled by money. As soon as these basic needs are satisfied, individuals are expected to focus on “higher” needs, such as love, esteem and self-actualization. Maslow hypothesized that these higher-order needs are rather to be met through engagement in considerable work than through monetary rewards (compare Rynes et al., 2005, p.574).

Herzberg’s motivation-hygiene theory (1957) focuses on classifying factors that contribute to either satisfaction, or dissatisfaction at work. For this, the factors involved in producing job satisfaction (and motivation) are separated and diverse from the factors that lead to job dissatisfaction. He differentiates between Motivators (e.g. challenging work, recognition) and Hygiene factors (e.g. job security, salary). Motivation factors are needed in order to motivate an employee to increase her performance actions. To ensure an employee not to be dissatisfied, Hygiene factors are needed. Herzberg saw hygienic needs as being driven by people’s “animal nature” (p.9) and posited that money was more likely to be a “hygienic” factor than a satisfying or motivating one. Thus, Herzberg believed that money played a role in creating or reducing dissatisfaction, but not in contributing to satisfaction or motivation (compare Rynes et al., 2005, p.574).

Deci and Ryan (1985) argue that when effort is applied in exchange for pay, pay takes on a controlling aspect that threatens the individual’s need for self-determination. Placing importance on monetary rewards is likely to decrease people’s intrinsic interest (for instance interest in the work itself), consequently dampening a potentially powerful alternative source of motivation. Deci and Ryan describe intrinsic motivation as “based in the innate, organismic needs for competence and self-determination” (p.33), and argue that it occurs in its purest form when “a person does an activity in the absence of a reward contingency or control” (p.35) (compare Rynes et al., 2005, p.575). This theory, combines different motivation factors and contributed largely to impact on follow-up analysis by economical research and is illustrated in greater details below.

Self Determination Theory

The Self Determination Theory (SDT) (Deci and Ryan, 1985) focuses on the individual needs for competence and self-determination. From this perspective, a person's intrinsic motivation is affected by changes in feelings of competence and self-determination. Linked with the SDT are the Organismic Integration Theory (OIT) and the Cognitive Evaluation Theory (CET). This literature claims, that the introduction of monetary rewards decreases task-specific intrinsic motivation under identifiable conditions. One consequence of the crowding-out of task-specific intrinsic motivation is that monetary rewards for performing a task may decrease the effort that is put onto the task.

Within SDT, Deci and Ryan introduced a subtheory, described OIT, to specify the different forms of extrinsic motivation and the related factors that either promote or hinder internalization and combination of the regulation for these behaviors. Figure 2 shows the different ways of motivation due to the OIT.

illustration not visible in this excerpt

Figure 2: Structure of Organismic Integration Theory[9]

The definition of OIT separates Motivation in three different groups, namely Intrinsic Motivation, Extrinsic Motivation and Amotivation. The Extrinsic Motivation is further subdivided in four divisions. They are specified as External, Introjected, Identified and Integrated regulation. The dimension of extrinsic motivation and corresponding intrinsic motivation lead to overall motivation of an individual and consequently to the effort provided on a task. Intrinsic motivation is increasing from External to Integrated regulation. Depending on the degree of self-determination and the amount of external control, one can categorize the effect in one of the four named subgroups.

Deci (1975, p.42) argues that “every reward (including feedback) has two aspects, a controlling aspect and an informational aspect which provides the recipient with information about his competence and self- determination.” This interpretation leads to the conclusion that offering an insufficient design of incentives might ends in a counterproductive motivation decrease and by this to a declining performance of the employee.

Further consideration is given to the reasons for intrinsic motivation which are described by the CET. This subtheory is providing relevant information about the individual motivation for performing a task out of an inner drive. The “CET framework suggests that social environments can facilitate or forestall intrinsic motivation by supporting versus thwarting people's innate psychological needs” (Ryan and Deci, 2000, p.71).

2.2 Economic Theory

In economics the focus on research dealing with the problem set of analyzing impacts of financial incentives on individual performance has been relying on the theoretical construction of the problem. Empirical results are in consequence of few available and difficult to survey data not many at hand.

This subchapter provides an introduction into the theoretical background of economical research on the way of functioning of financial incentives and individuals’ behavior.

The Principal-Agent-Theory deals with the relationship of employer (principal) and employee (agent) and in which way contracts and incentives between those two groups should be modeled for reaching an optimum for both sides.

An approach to analyze the individual motivation differences between the absence of financial incentives and the implementation of such is the Motivation Crowding Theory (Frey, 1997; Frey and Oberholzer-Gee, 1997; Bénabou and Tirole, 2003). This takes into account the existence of individuals’ intrinsic motivation and attempts to explain how motivation is impacted by the application of financial incentives.

Solutions to the motivation crowding problem are suggested through the modeling of optimal contracts. One approach is including this phenomenon into the optimization constraint by regarding the relation between principal and agent from a dynamic perspective (Murdock, 2002).

Beside the Motivation Crowding Theory, it is questionable how individual motivation is affected as incentives further increase. The effects of differing wages on individual motivation to work, and continuing on the relationship between leisure and working time is the subject of the third subpart of this subchapter where the constructs of reference dependent preferences and the backward-bending labor supply curve are introduced.


The economic theory dealing with interactions between individuals is described as the principle-agent theory. Part of this research area is the studying of optimal contracts, property rights and organizational theory. A short and obvious definition of the subject was given by Pratt and Zeckhauser (1985) as they formulized the problem as following: “Whenever one individual depends on the action of another, an agency relationship arises. The individual taking the action is called the agent. The affected party is the principal.” As a consequence of asymmetric information, problems in providing an optimal contract arise. The information problems are defined as Moral Hazard, Adverse Selection, Hidden Information, Multiple parties and non-verifiability versus non-observability[10]. According to the principal-agent paradigm and being faced with asymmetric information, firms should attach the salary of employees to any verifiable (individual or collective) signal of performance.

Aligning the interests of the agent in solidarity with those of the principal, various mechanisms can be used. Two ways to manage moral hazard problems are pointed out by the classical agency theory. The first one is monitoring of the agents’ behavior and the second one is guaranteeing an optimal performance by applying incentive contracts[11].

For economical research it has been obvious for long times that “raising monetary incentives increases [work] supply” (Frey and Jegen, 2001, p.590) and that “a well-established result of most standard hidden action models is that higher incentives, ceteris paribus, lead to higher performance” (Pokorny, 2006, p.251). In other words: according to theory, paying higher incentives by the principal has to lead to higher motivation and effort by the agent and through this to a higher performance. Caused by this elementary economical law, research focused on the relationship between extrinsic motivation (especially rewards) and performance or effort. At the same time, intrinsic motivation has not been considered as impacting overall performance and therefore, it was seen as a right-shift of the supply curve as an exogenously amount of work is added (Frey, 1997, p.428). Newer theoretical perspectives include intrinsic motivation into consideration and model extrinsic motivation effects (e.g. Frey, 1997; Frey and Oberholzer-Gee, 1997; Bénabou and Tirole, 2003).

Motivation Crowding Theory

As solely extrinsic motivation has been relevant for economists to define the optimization problems of a principal-agent contract, the style of optimal incentive-compatible contracts disregarded the factor of intrinsic motivation. This perception was partly confuted by the Motivation Crowding Theory, which generates the link between intrinsic motivation and Agency Theory (Herpen et al., 2005, p.307).

The implication of this is a possible intrinsic motivation crowding-out due to improper extrinsic motivation which is followed by a decreasing effort, although the amount of incentives has been raised. An illustration of the crowding-out effect and the price effect on individual effort in consequence of increasing incentives can be found in Figure 3, where those effects are illustrated and the interference on the overall effort an agent spends is drawn.

illustration not visible in this excerpt

Figure 3: Illustration of the Crowding-Out Effect of Intrinsic Motivation[12]

In economic theory, two major reasons for the decreasing performance caused by the introduction of extrinsic motivation factors have been identified. First, a change in preferences and second, a change in the perceived nature of the performed task, in the task-environment or in the actor’s self-perception (compare Frey and Jegen, 2001, p.592).

The theory of changes in preferences was introduced by Frey (1997), who combined the psychological and the economical view on individual performance and managed (as one of the first economists) to bring both disciplines together. The intention is the existence of a possibility for an individual’s preference change, which is followed by a motivation crowding-out or likewise a motivation crowding-in. Frey states, that this effect can appear in every task where intrinsically motivation is present. As a result “the relative price effect of monetary compensation is not questioned in any way, but this measure becomes less effective when crowding-out is considered” (Frey and Oberholzer-Gee, 1997, p.754).

He identified two reasons for the occurrence of motivation crowding. The first one is a high intrinsic motivation by the agent before an extrinsic motivation factor is applied, and the second one is the presence of the conditions for crowding-out of intrinsic work motivation (Frey, 1997, p.431). The high work moral (or intrinsic motivation) can for instance be achieved through interesting tasks, a personal relationship between the principal and the agent or/and a higher participation possibility of the agent (p.431). Frey identified four conditions which systematically influence the way to perceive external interventions by the agent (p.432). First, differing effects of rewards and commands: The agents feel a certain amount of freedom in their intensity of responding through the application of rewards, as agents might appreciate rewards to be less restrictive to their self-determination than commands. Second, the more a reward is contingent on the performance, the more strongly the locus of control is shifted from intrinsic to extrinsic incentives and the more intrinsic motivation is crowded out. Third, dependent on the extent of differentiation made between all agents, an agent perceives the principals intervention to be controlling or informing. Fourth, the more strongly an external intervention implies an acknowledgement of the agent’s high intrinsic motivation, the more strongly it fosters work moral.

Bénabou and Tirole (2003) showed the hidden costs of rewards and as a result the crowding-out of intrinsic motivation. The change in agents point of view leads to the result that “Incentives are […] only weak reinforces in the short run, and negative reinforces in the long run” (p.489) and consequently produce hidden costs in the length of time, and by this lead to a motivation crowding-out. The possibility of a motivation crowding-in or in other words an increase of overall performance due to higher extrinsic motivation explicitly exists.

For Bénabou and Tirole major cases of intrinsic motivation are the task attractiveness and trust effect. Two major negative impacts on self-confidence are identified for rewards. The first one is information, as the principal has information about the agents’ everyday jobs. In most job positions the rewards for all workers with comparable jobs are the same or at least similar. The second one is the sorting condition. In equilibrium “rewards undermine the agent´s assessment of the task´s attractiveness” (p.497). They point out that a higher reward and a lower base salary is “always bad news for the agent permanently damages his self-confidence, no matter what the task´s outcome turns out to be” (p.501).

If the agent has the same information about a task as the principal, the reward could have a positive impact on motivation. Otherwise, if the agent is only receiving a signal by adjusting incentives, this could undermine the agent´s expectation as she updates her belief´s using the principal’s equilibrium strategy. The task attractiveness is reduced through an increase of the reward. Especially for minor ability jobs a more performance based reward should be offered, as those payments improve the motivation of the workers. By this modeling of a contract, the intrinsic motivation of an agent could be undermined by extrinsic motivation adaption.

Empirically, motivation crowding is difficult to observe as data are almost impossible to survey. Gneezy and Rustichini (2000b) analyzed the impact of a financial penalty for a specific behavior in a field study. The study was conducted in 11 private day-care centers in the city of Haifa, from January to June 1998 (20 weeks). In the first 4 weeks the number of parents who arrived late was observed. At the beginning of the fifth week, a small fine was introduced in 7 out of the 11 day-care centers. The other 4 day-care centers served as a control group and no fine was introduced. The fine was imposed on parents who came more than 10 minutes too late.

The result of introducing the fine was a sharp and significant increase in the number of parents coming late (p.3). At the beginning of week number 17 the fine was removed, but in the following 4 weeks this had no effect on the parents coming late. The authors see in this field study by the introduction of fines a change in the perception of people regarding the environment in which they operate. The environment in the study has been defined by an incomplete contract. The introduction of the fine into this incomplete contract reshaped the perception of the parents regarding this environment. By this, the fine had a similar effectiveness as financial incentives. It yielded to a change in preferences and a crowding-out of intrinsic motivation for delivering the children in time as a result of the application of financial incentives.

Research on the Design of Optimal Contracts

Further research on the intrinsic motivation crowding-out due to extrinsic motivation has been conducted. One goal of this research was the approach of modeling optimal contracts for taking into account intrinsic motivation and minimizing motivation crowding-out even though extrinsic motivation tools are applied.

Based on the aspects of Motivation Crowding Theory, Murdock (2002) analyzed the role of optimal contracts regarding intrinsic motivation. He states that “contracts are necessarily incomplete when agents are intrinsically motivated” (p.651). As a result he was showing implicit contracts and intrinsic motivation to be complements and performance has to be considered in a dynamic environment. Furthermore, he suggested regarding the status of implicit contracts, companies to sometimes implement projects with a negative expected profit but with large intrinsic returns to the agent (p.667). Through this, the expected profit through each project might be lower, but this helps to enhance private benefits of the agent and satisfies employees as the intrinsic motivation in those projects is especially high. The agent, however, works harder, and this increased effort generates more positive-expected-profit projects for the firm. This increased rate of project generation more than compensates for the lower profit per project (p.667). Caused by higher intrinsic motivation in comparison to normal sense, Murdock suggests that the increased effort yields to a creation of more positive-expected-profit projects for the firm (in numbers) and by this to an overcompensation of the lost lower profit per project. In this calculation of the intrinsic motivation as a fundamental part of the contract and by worker´s satisfaction with the assigned projects, the overall profits for the principal’s respective company may rise.

Incentives and Reference Dependent Preferences

There is an ongoing controversy whether the response of labor supply to changes in incentives is consistent with the standard intertemporal substitution of labor and leisure, or rather with loss aversion around a daily reference income (see Abeler et al., 2009, p.3). The theoretical approaches are introduced below to provide an insight on both views.

The impact of increasing wages on working motivation is described by the backward-bending labor supply curve which is illustrated in Figure 4 (see also Gilbert and Pfouts, 1958; Hanoch, 1965).

In this approach, the relation between increasing wages and the willingness for spending additional hours on income enhancement is illustrated (compare Figure 4).

illustration not visible in this excerpt

Figure 4: Illustration of the Backward Bending labor Supply Curve (own illustration)

As wage increases, people substitute leisure for working time. Based on this, as the income increases, the performance is supposed to decrease as soon as the level of income reached a certain extent (in the graph marked as point P). If wages rise up to this point, the worker obtains a higher expected utility from more income than from more leisure time. Consequently she is willing to increase her hours worked. Those hours may be hours worked per day, month, year or even lifetime. Within this section of the curve the substitution effect is positive while the income effect is negative. However, as the substitution effect is superior to the income effect, this is giving rise to a positive price effect. Therefore, the increase in the real wage rate will cause an increase in the number of hours worked.

If the real wage increases further beyond P, the number of hours worked would fall as the income effect surpasses the substitution effect. This is because of a greater utility gain from an extra hour of leisure than the utility gained from the additional income earned through working. Basically, beyond the wage of P the worker is being paid enough to sustain the current lifestyle without having to work more hours. Therefore the result is a backward bend in the curve. This theoretical approach is based on the hours worked and not on tasks.

Moving along, since people might make additions to their actual wage level, evaluate their utility and take a reference level (e.g. an expected wage or a rival’s/workmate’s wage) into account, the hypothesis of a reference level of compensation seems to be straightforward. The relevance of reference points has been the objective of research in many fields, and a growing class of theories (e.g. Kahneman and Tversky, 1979; Tversky and Kahneman, 1991; Bell, 1985; Loomes and Sugden, 1986; Gul, 1991; Shalev, 2000; Köszegi and Rabin, 2006, 2007, 2009) are built on the idea that expectations can act as a reference point and these models are able to align empirical evidence that is hard to reconcile with usual economic assumptions.

By analyzing decisions under uncertainty Kahneman and Tversky (1979) introduced the theory of reference points (compare Pokorny, 2006, p.259). For this, they developed a model describing loss aversion by modeling a utility function including a reference point. This utility function has a convex slope below the reference point and develops a concave slope above the reference point. Hence subjects behave risk seeking below the reference point and risk averse above it. Tversky and Kahneman (1991) expanded the application of reference dependent utility by suggesting that no additional effort is rational after reaching the reference level, as costs would exceed utility gains from wage. The utility function increases linear in wage but develops a smaller slope as a reference wage is reached. The reference wage is defined as a point from where wages are evaluated. This might be a wage the agent expects or perceives to be appropriate for a certain task. After that point is reached the slope of the utility function flattens. Thus, utility increases more slowly if wages exceed the reference point.

In an analysis of the behavior of New York cab drivers, Camerer et al. (1997) investigated the relationship between wages and hours worked. For this, they analyzed three different data samples of cab drivers in New York City. In total, 1826 samples throughout 1988 and the years between 1990 and 1994 had been considered for the study.

As a result of their analysis, they point out that the cab drivers, after having once reached a certain amount of income during a day, reduce their effort on further earnings by stopping or limiting their input on more rides. In other words: if their hourly earnings within a day were higher, they worked altogether less, and if their hourly earnings were lower, they went on working until a specific level of income has been reached. The authors assume that the cab drivers set a target point of income each day and as soon as this is reached, their motivation and willingness to put further effort onto additional earnings decreased and leisure time became more important than a higher income (p.432). The cab drivers conduct an intraday optimization, but do not optimize among the days. Those assumptions are encouraged as numerous drivers’ even quit work, though not having reached the end of the 12 hours shift because they gained the targeted income.

Support for the existence of a rational expectation about a wage was provided by subsequent research (e.g. Chou, 2002; Goette et al., 2004; Farber, 2005; Fehr and Goette, 2007; Farber, 2008; Crawford and Meng, 2008). Those studies furthermore found evidence for an estimated income and supply of work targeting around this expectation.


[1] For further reading on the financial crisis of 2007 refer to the literature (for instance Ambachtshee et al. (2008)) or to the internet.

[2] Nevertheless, studies analyzing the implementation of performance-based payments in comparison to flat-payments in companies are on hand and presented in the course of this thesis.

[3] For an introduction into the development of Experimental Economics and for a discussion on advantages and disadvantages of this discipline please refer to the literature (for instance Friedman and Sunder (1994) or Kagel and Roth (1997)).

[4] adapted from Bonner and Sprinkle, 2002, Fig.1

[5] The meta-analysis tool is a relatively new statistical method that allows researchers to summarize the results of many different experiments conducted on a specific topic by different scientists at different points in time by estimating the effect size of treatments such as incentive systems as a percentage of a standard deviation change in performance due to the strategy being investigated (see Condly et al., 2003, p.46).

[6] This can be especially harmful for tasks involving insight or creativity (both require a kind of open-minded thinking that enables one to draw unusual connections between elements)

[7] Compare Langer and Imber, 1979; Camerer et al., 2005

[8] For further reading please refer to the psychological literature on motivation (for instance Miner, 2005; Latham, 2006). Considerable motivation theories for further reading are especially: Vroom (1964), Atkinson (1957), Weiner (1986), Festinger (1978), Hull (1943), Csikszentmihalyi (1975), Locke and Latham (1990).

[9] own illustration according to Gagné and Deci, 2005, p.336, Fig. 1

[10] For a further overview and details refer to the literature (for instance every standard economic textbook on microeconomics)

[11] For this thesis, the knowledge about the modeling of optimal contracts is too extensive and also the application of mathematical models and optimization constraints. For further reading refer to standard economical textbooks or the literature which is mainly dealing with the theoretical aspects of contracts (for instance Mirrlees, 1976; Grossmann and Hart, 1983; Holmstrom and Milgrom, 1987; Hart and Holmstrom, 1987; Milgrom and Roberts, 1992).

[12] adapted from Weibel et al. (2007)


ISBN (eBook)
805 KB
Institution / Hochschule
Universität Mannheim – Volkswirtschaftslehre
financial incentives experiment finanzielle anreize experimentelle wirtschaftsforschung principal agent



Titel: The Impact of Financial Incentives on Individual Performance: An Experimental Approach