An Experimental Investigation of Employer Discretion in Employee Performance Evaluation and Compensa

THE ACCOUNTING REVIEWV ol.80,No.22005pp.563–583An Experimental Investigationof Employer Discretion inEmployee Performance Evaluationand CompensationJoseph G.FisherIndiana UniversityLaureen A.MainesIndiana UniversitySean A.PefferUniversity of KentuckyGeoffrey B.SprinkleIndiana UniversityABSTRACT:Employment relationships provide fertile ground for both employee andemployer opportunism.Employers worry about whether employees will devote suffi-cient effort to work,and employees are concerned about whether employers will com-pensate them appropriately.In this paper,we examine whether employer discretionover the size of the total employee compensation pool and the allocation of this poolamong employees influences employee and employer opportunism.The results of ourexperiment indicate thatfirm output and employees’compensation are greater whenthe employer does not have discretion over total employee compensation,but doeshave discretion over the allocation of total compensation.Wefind that the employer’sresidual profit increases with discretion over the allocation of compensation amongemployees;however,wefind no effect on residual profit of the employer’s discretionover the total amount of employee compensation.Our results suggest thatfirms benefitfrom a compensation contract that establishes total employee compensation as a pre-determined function of public,aggregate measures such as accounting income,butWe thank Walt Blacconiere,Dave Barrett,Daniel Beneish,Bob Bowen,Dave Greene,John Hassell,Dan Heitger, Frank Hodge,Jane Jollineau Kennedy,Jamie Pratt,Jerry Salamon,Terry Shevlin,Reed Smith,Michael Williamson, two anonymous reviewers,and participants at research workshops at Emory University,Indiana University Bloom-ington,and Indiana University Indianapolis,University of Maryland,Michigan State University,University of Missouri,The University of North Carolina at Chapel Hill,University of South Carolina,and University of Wash-ington for helpful comments and suggestions.The University of Kentucky and the Pressler Faculty Fellowship at Indiana University provided generous funding for this project.Editor’s note:This paper was accepted by Marlys Gascho Lipe,Editor.Submitted July2003Accepted Sept2004563564Fisher,Maines,Peffer,and SprinkleThe Accounting Review,April 2005provides the employer at least some discretion to allocate this compensation using private information.However,our results caution that employees and employers may not have similar preferences for the degree of employer discretion over the determi-nation of total employee compensation.Keywords:compensation;discretion;subjective performance evaluation;double-sidedmoral hazard.Data Availability:Contact the authors.I.INTRODUCTIONEmployment relationships typically are characterized by two-sided concerns regarding opportunistic behavior (Demski 1997,579).Employers are concerned about whether employees will work diligently,while employees worry about whether employers will reward this work appropriately.The potential for employees and employers to act in their own self-interest when making these choices threatens to destroy the cooperation needed for productive firm outcomes.Opportunism may not arise if workplace norms result in cooperation among employers and employees.However,firms typically do not rely solely on workplace norms to pre-empt opportunism,and implement control systems,such as performance evaluation and compensation systems,to mitigate opportunism.These systems govern two compensation decisions:(1)how to allocate total remuneration between the employer and employees and(2)how to allocate total employee remuneration among employees (Baiman and Rajan 1995).The first decision establishes the size of the total employee compensation pool and the second determines the allocation of this pool among employees.The approach a firm uses to make these decisions can influence both employer and employee opportunism.In this paper,we examine whether an employer’s discretion over the two compensation decisions affects employee and employer opportunism.The effect of employer discretion over the two compensation decisions is unclear because discretion creates tension with respect to employer and employee opportunism.An employer can use discretion to her personal benefit,increasing her share of total compensation at the expense of employees or allocating compensation to ‘‘favored’’employees.1However,discretion also allows an employer to reduce employee opportunism by determining employee compensation using measures and approaches that best reflect employees’collective and individual efforts.The tension between employer discretion and employer/employee opportunism typi-cally is connected to firms’information systems.Firms’formal information systems,such as accounting systems,provide information that is publicly observable or verifiable but often reflects only aggregate employee performance (Baiman and Rajan 1995).Employers can alleviate employees’concerns about employer opportunism by giving up discretion and committing ex ante to a total compensation pool and its allocation among employees based on such public information.However,public information may not reflect all relevant aspects of performance.In particular,aggregate information provides little basis for employers to differentially reward individual employee performance and prevent employees from free-riding on the efforts of other employees.Thus,compensation approaches based on publicly available information may reduce employer opportunism but may not prevent employee opportunism.1Throughout the paper,we use feminine pronouns to refer to employers and masculine pronouns to refer to employees.An Experimental Investigation of Employer Discretion565 Firms also have informal sources of information.This information can reflect multiple aspects of aggregate performance and/or isolate the performance of individual employees, but is privately observed by the employer.An employer can reduce employees’opportunism by retaining discretion to use her private information in performance evaluation.For ex-ample,an employer can reduce the benefit of free-riding by using her private information on individual performance to allocate the compensation pool among employees.Unfortu-nately,allowing the employer discretion to use private information generates employee concerns that the employer will fail to reward employees appropriately for their aggregate or individual efforts.Thus,a compensation scheme in which an employer has full discretion to use private information may reduce employee opportunism but allows for employer opportunism.The relative strengths and weaknesses of restricting versus allowing employer discretion suggests that there may be complementary roles for discretion versus nondiscretion with respect to the two compensation decisions.Specifically,an approach in which the employer does not have discretion over the size of the total employee compensation pool(hereafter pool size)but does have discretion over the allocation of the employee compensation pool (hereafter pool allocation)may alleviate concerns about both employer and employee op-portunism(Baiman and Rajan1995).This approach suggests that there also are comple-mentary roles for public and private information—public information provides the basis for the total compensation pool and private information assists in the allocation of this pool.To investigate the effects of employer discretion on employee and employer opportun-ism,we examine situations in which the employer either has full discretion or no discretion over the size of the employee compensation pool and/or the allocation of this pool among employees.We use a laboratory experiment with a2(pool size discretion)ϫ2(pool allocation discretion)ϫ8(period)mixed design.In the experiment,the employer either has discretion to determine the size of the two-employee compensation pool(discretionary pool size,DS)or the pool is set at two-thirds of the revenue from total employee output (nondiscretionary pool size,NDS).Similarly,the employer either has discretion in allocat-ing the compensation pool between the two employees(discretionary pool allocation,DA) or the pool is split evenly between the two employees(nondiscretionary pool allocation, NDA).The employer receives the residual profit after subtracting employee compensation from total revenue.In the experiment,237undergraduate business students,in groups of three,acted either as an employer or as one of two employees.Employees allocated three minutes between a work task(decoding numbers into letters)and compensated leisure for each of eight periods. After each work session,employers in the discretionary conditions determined the size of the employee compensation pool and/or the allocation of this pool between the two em-ployees.Employers received noisy private information about individual employee output to use if desired in making their compensation decisions.Thus,in the discretionary conditions, employers had the opportunity to more closely align an employee’s compensation with individual production than in the nondiscretionary conditions in which compensation was a multiple of total output and/or evenly split between the two employees.We consider the impact of employer discretion over pool size and pool allocation on total output,as well as the division of the value of total output(i.e.,revenue)between employer remuneration and employee remuneration.The results indicate that both total output and employee compensation benefit from restricting the employer’s discretion over the size of the compensation pool,and allowing the employer to have discretion over the allocation of this pool.Total output and employee compensation were greater when theThe Accounting Review,April2005566Fisher,Maines,Peffer,and Sprinkle employer did not have discretion over pool size than when she did have discretion,as well when the employer had discretion over the pool allocation than when employees received equal compensation.These results hold even when we eliminate the last two periods to control for possible end-game effects.Our results for the effect of discretion on the employer’s residual profit are mixed.Over the entire eight periods,the employer’s residual profit did not differ between the discre-tionary and nondiscretionary pool size conditions.The employer’s residual profit was greater in the discretionary pool allocation condition than in the nondiscretionary condition. However,we found no effect of either pool size or pool allocation discretion on employer’s residual profit after eliminating thefinal two periods.This result suggests that discretionary pool allocation primarily mitigates employee opportunism that arises as the employment relationship nears its end.Our results also document a fairly high level of cooperative behavior between employ-ers and employees in the absence of controls that mitigate opportunism.In the DS-NDA contract,which provides no explicit controls to mitigate employee or employer opportun-ism,employees produced,on average,75percent of the amount produced in the best-performing NDS-DA contract.Additionally,employers with discretion over pool size paid out,on average,59percent of revenues as compensation to the two employees.However, the effects of opportunism associated with either a discretionary pool size or nondiscre-tionary pool allocation approach increased over the eight periods.For example,in the last period,output in the DS-NDA contract was only45percent of that in the NDS-DA contract. Moreover,employers in the two discretionary size conditions paid only51percent of rev-enue to employees in the last period.This study contributes to the control systems literature by examining implications of employer discretion in performance evaluation and compensation for employer and em-ployee opportunism.Our results document a fairly high level of productive output in the absence of specific controls related to employer discretion/nondiscretion intended to miti-gate employee and employer opportunism.Rather,employers and employees appear to recognize the economic benefits of joint cooperation and rely on social values,such as trust and reciprocity,to keep productive outcomes intact over all but thefinal periods of em-ployment(Axelrod1984;Berg et al.1995;Sprinkle2003).Thesefindings highlight that the efficacy and cost of performance evaluation and compensation arrangements should be compared not simply against each other,but against the cost and efficacy of naturally occurring social mechanisms in the workplace.Despite these cooperative outcomes,our study documents benefits of control systems with explicit incentives to mitigate employer and employee opportunism.Specifically,al-lowing the employer discretion to allocate compensation among employees increasesfirm output and both employee and employer remuneration.However,the implications of our results for discretion in setting the size of total employee compensation are less clear.While eliminating employer discretion over the pool size improves output and employee compen-sation,the employer’s residual profit is not greater under the nondiscretionary pool size case than under the discretionary pool size case.In essence,the benefit to the employer of increased output associated with a nondiscretionary pool size is not sufficiently large to offset the cost to the employer of having to pay greater per unit compensation to employees in our experiment.Thus,employers may prefer to retain discretion over both pool size and pool allocation,to the detriment offirm output and employee compensation.However,given the compensation-related benefits to employees of a nondiscretionary approach to pool size, competition among employers for employees could lead employers to relinquish their dis-cretion over compensation pool size.The Accounting Review,April2005An Experimental Investigation of Employer Discretion 567The Accounting Review,April 2005Finally,our study complements research on the use of contractible (public)versus noncontractible (private)information in compensation contracts.Analytical research docu-ments the benefits of compensation contracts that base the total pool size on contractible information and the pool allocation on employer’s private information (Baiman and Rajan 1995).Our results generally provide empirical support for Baiman and Rajan’s (1995)conclusion.Since contractible information often emanates from formal accounting systems,our findings also suggest that there are complementary roles in designing compensation systems for information from formal accounting systems and private information from a firm’s informal information systems.The remainder of this paper is organized into four sections.The next section provides background and develops the hypotheses,and Section III describes the experiment used to test these hypotheses.Section IV presents the results,and Section V provides a summary and discussion of the results.II.BACKGROUND AND HYPOTHESESEmployer and Employee Opportunism—Discretionary Pool Size/Nondiscretionary Pool Allocation Contract (DS-NDA)We examine a setting in which employees work independently on a production task for a finite number of periods.Each employee allocates a fixed amount of time each period between work and leisure,and derives utility from both leisure time and compensation.Output for the group of employees is the sum of individual employee output.Firm revenue equals group output multiplied by the competitive market price.The employer is a residual claimant who receives total firm revenue less the amount paid as compensation to employees.The firm has an information system that measures employee performance and a com-pensation system that rewards employee performance (Zimmerman 2003,169).To study our research question on the effects of employer discretion on opportunism,we vary char-acteristics of these two systems.We first examine a setting that provides no formal controls to mitigate either employer or employee opportunism and then add controls that potentially mitigate one or both types of opportunism.In our initial setting,the information system measures group output (e.g.,revenue or income for the entire company or a division),but not individual employee output.2The employer retains discretion to determine the size of employees’compensation pool.Given the lack of information on individual performance,the employer does not retain discretion over the allocation of the compensation pool.Rather,the pool is divided equally among the employees.We term this compensation system the discretionary size/nondiscretionary allocation (DS-NDA)contract.These choices for information and compensation systems represent company settings that provide fertile ground for both employer and employee opportunism.Employees have an incentive to act opportunistically and shirk in their work effort because their individual compensation is based on group performance.If an employee receives more utility from leisure than from the expected incremental compensation he would receive from working,2In practice,employers often observe aggregate performance measures from firms’formal information systems,but can not disentangle individual employee performance due to complexities of the production process.In these settings,employers frequently use aggregated performance as a basis for incentive contracts and allocate total compensation in a manner unrelated to individual performance (Milgrom and Roberts 1992,413).568Fisher,Maines,Peffer,and Sprinkle The Accounting Review,April 2005then he may shirk and receive work compensation based on the efforts of his co-workers (free-ride).3Employers also have an incentive to act opportunistically given their discretion over the size of total employee compensation.Given a finite-period world,a self-interested em-ployer will pay no employee compensation in the last period since it only reduces her residual payment and has no incentive value for employees’future work-leisure allocation decisions (employees have already made all such decisions).If employees believe the em-ployer will not pay compensation in the final period,then they have no incentive to work and will allocate all time to leisure.Backward induction implies a Nash equilibrium in which the employer does not pay compensation and employees do not work in any period.4Both employer and employee opportunism threaten to destroy productive outcomes that could benefit all firm members.This threat arises from employer discretion over total em-ployee compensation and employee discretion over work effort,combined with incentives for employers and employees to use their discretion for personal gain.The threat of opportunism may not materialize if employees and employers choose to cooperate with one another,either out of recognition of their mutual benefit or out of fear that others will retaliate against opportunistic behavior (Fehr and Ga ¨chter 2000).Such cooperative behavior requires trust that others will not act opportunistically.Research on trust games shows that individuals are willing to place themselves at risk by cooperating and trusting that other individuals will not exhibit self-interested behavior (Berg et al.1995).If all employees initially trust that other employees will choose to work rather than to shirk,then a productive equilibrium can be sustained as long as all employees reciprocate.Indeed,Fisher et al.(2003)find that employees’work allocations and performance are significantly greater than zero in a setting that does not provide a compensation-based mechanism to mitigate employee opportunism.5Employees also must trust that the employer will not act in her own self-interest,but instead will reciprocate by compensating employees appropriately.This expectation is not without foundation,as research indicates that employers reward employees who exert high levels of effort (Fehr et al.1997).Additionally,over time,an employer can develop a reputation for compensating employees that induces employees to work in advance of re-ceiving compensation (Bull 1987).Thus,it is possible that extreme opportunism will not arise in the DS-NDA contract due to social values such as trust and reciprocity.However,if these social values are not sustainable workplace norms,then other mechanisms are needed to mitigate employee and employer opportunism.The firm’s performance evaluation and compensation system can serve as such a mechanism.In the next section,we use implications of a paper by Baiman and Rajan (1995)to consider how changes to the employer’s discretion in setting employee compensation can help to mitigate employer and employee opportunism.Specifically,we examine three compensation contracts that change one or both aspects of the DS-NDA 3This is similar to public goods settings in which individuals have the incentive not to contribute to the common pool,but consume the benefits of contributions from other individuals (Ledyard 1995).4For example,the employer knows that paying compensation in the next-to-last period cannot motivate employees’effort for that period since production has already occurred,but could motivate employees’effort in the last period.However,the employer knows that employees will not work in the last period given that she can not commit to paying compensation.Thus,there is no incentive for her to pay compensation in the next-to-the-last period.5However,Fisher et al.(2003)include a production goal in their setting,which may act as a separate mechanism influencing employee behavior.An Experimental Investigation of Employer Discretion 569The Accounting Review,April 2005contract with respect to employer discretion/nondiscretion in determining the size of the employee compensation pool and/or the allocation of this pool among pensation Contracts to Mitigate Employer and Employee OpportunismBaiman and Rajan (1995)provide insights on features of compensation contracts that have potential to mitigate employee and employer opportunism.Baiman and Rajan (1995)(hereafter B-R)examine a contract in which the employer commits ex ante to determine the size of the employee compensation pool based on jointly observable (contractible)information,but does not commit to a specific pool allocation.Rather,the employer has access to private (noncontractible)information about an individual employee’s performance and retains the right to use this information to allocate the compensation pool.B-R ana-lytically show that this contract results in a strict Pareto improvement over a compensation contract in which the employer commits to determining compensation using only jointly observable information.The findings in B-R highlight the importance of employer discretion for employee and employer opportunism associated with the DS-NDA contract.The optimal contract in B-R has two features that differ from the DS-NDA contract.First,in the B-R contract,the employer gives up discretion to set the compensation pool size ex post .Second,the em-ployer retains discretion to allocate the compensation pool among employees using her private information related to individual employee performance.Thus,B-R’s analytical results suggest that employee and employer opportunism can be mitigated by choices related to employer discretion.We discuss these issues below for employee and employer opportunism.Mitigating Employee Opportunism—Discretionary Pool Size/Discretionary Pool Allocation Contract (DS-DA)In the DS-DA contract,the employer has complete discretion over both the amount of the compensation pool and the allocation of the pool among employees.While the DS-DA contract does not alleviate employer opportunism due to this discretion,the contract pro-vides an opportunity for mitigating employee opportunism.Specifically,if the information system is expanded to include the employer’s private noisy signals about each employee’s output (as in B-R),the employer can use her discretion to determine the allocation of compensation among employees based on private information about individual employee performance.6By more closely aligning an employee’s compensation with his individual performance,the employer reduces the benefits of shirking to the employee,which should increase his work efforts.Mitigating Employer Opportunism—Nondiscretionary Pool Size/Nondiscretionary Pool Allocation Contract (NDS-NDA)In this contract,the employer has no discretion in setting either the size of the com-pensation pool or its allocation among employees;in essence,the employer has no role in determining employee compensation.The total employee compensation pool is a function of output (percentage of total revenue)and employees receive an equal share of the com-pensation pool.The equal division of the compensation pool among employees provides 6The characterization of the employer’s private information as a noisy signal of individual employee performance reflects the idea that individual employee output typically is not easily identified (Prendergast 1999,11).How-ever,employers often receive information related to aspects of individual employee performance.This infor-mation typically is private and based on subjective assessments.If the signals were public rather than private,then an employer could explicitly tie an employee’s compensation to his individual performance signal.570Fisher,Maines,Peffer,and Sprinkle The Accounting Review,April 2005no formal controls for employee opportunism;however,the NDS-NDA contract does mit-igate employer opportunism by taking away the employer’s discretion.Note that the NDS-NDA contract is more extreme in this respect than the optimal B-R contract.The employer has no discretion in the NDS-NDA contract,while the employer in the B-R contract retains discretion over the parameters of the compensation pool but commits ex ante to these parameters using jointly observable information.Thus,the NDS-NDA contract focuses on eliminating employer opportunism,as opposed to creating an optimal contract from the employer’s viewpoint,as in B-R.7Mitigating Both Employee and Employer Opportunism—Nondiscretionary Pool Size/Discretionary Pool Allocation Contract (NDS-DA)In the NDS-DA contract,the compensation pool is a predetermined function of group output,but the employer has discretion in allocating the pool among employees.Employer opportunism is mitigated because the employer has no discretion over the size of the com-pensation pool.Moreover,the employer can reduce employee opportunism by using her private signals about individual performance to align an employee’s compensation with his performance.Thus,the NDS-DA contract has potential to mitigate both employer and employee opportunism.HypothesesThe prior discussion indicates that use of a nondiscretionary approach to determining the size of the employee compensation pool can mitigate employer opportunism.This in-creases employees’willingness to devote time to work,which,in turn,increases output.Similarly,the use of a discretionary approach to allocating the employee compensation pool should mitigate employee shirking and increase output.Employees will devote more time to work because they expect that their employer will exercise her discretion and use infor-mation about their individual production to determine their share of total employee com-pensation.Ceteris paribus ,both employees and employers should benefit from the increased output.These arguments lead to the following hypotheses.Employees’Group Output (Performance)8H1a:Group output will be greater when the employer does not have discretion overthe size of the employee compensation pool than when the employer has discre-tion over the size of the employee compensation pool.H1b:Group output will be greater when the employer has discretion over the allocationof the employee compensation pool than when the employer does not have dis-cretion over the allocation of the employee compensation pool.7There are several other differences between B-R and our paper.First,the B-R model is a single-period model while our setting includes a finitely repeated setting.Second,in the B-R model,outcomes and signals are dichotomous,while in our setting production outcomes and signals can take on numerous levels.Finally,in the B-R model,there are jointly observable signals about each employee’s individual performance,whereas in our setting all individual signals are the employer’s private information and only group output is jointly observable.Thus,while B-R has implications for our setting,our paper is not a direct test of their analytical predictions.8Predictions similar to H1a and H1b could be made for employees’allocation of time to work.Given our assumption that output is a function of work time,we provide hypotheses only for output (performance).We report empirical results for employees’work time in footnote 13.。