When the concept of management by objectives (MBO) became widely recognised in the 1950s, performance management systems moved towards those based on achieving goals, or success in key accountabilities (or key job impact areas)(Pratt 2001). Research between 1950 and 1980 was focussed primarily on improving the instruments used in making performance ratings, different types of rating scales, rating vs. ranking, and different methods of eliciting ratings that would provide the most objective measures of performance were investigated (Arvey Murphy 1998). Problems associated with goals and accountabilities appraisal programs are related to goals and objectives being primarily forecasts of what could be achieved and as a result, are affected by a variety of external influences often outside the control of the individual employee being assessed. Sometimes allowances were made for events such as budget changes, abrupt work rule changes and the like, but often subordinates were unfairly held responsible for matters outside their control. The other problem area lies in the negotiation of goals and achievements that will form the basis of subsequent performance evaluation and the employee's interest often lying in setting goals that are easy to both achieve and exceed in preference to ones that would require hard work to achieve (Pratt 2001).
Flaws inherent in most performance appraisal systems stem from the premise that "it is not clear that performance is a knowable and observable objective reality and performance ratings are reasonable reflections of that reality" (Burns 1996). Much recent performance evaluation research focuses on defining job performance (see Campbell 2003 Borman 2003). The domain that is job performance is expanding with a move towards more flexible definitions of work roles and jobs. Jobs are viewed as dynamic and more interchangeable and the focus is on personal competencies required to perform various work roles and jobs rather than a review of specific tasks and duties inherent in fixed jobs and work roles (Ilgen Hollenbeck 2001). Broad distinctions have been made between measures of task proficiency in job performance and what is being called 'contextual performance' (Arvey Murphy 1998). Borman Motowidlo (2003) define task performance as the proficiency with which incumbents perform core technical activities that are important for there jobs and contextual performance as extratask proficiency that contributes more to the organisational, social, and psychological environment to help accomplish organisational objectives. Contextual factors include such aspects as persisting with enthusiasm and extra effort, volunteering to carry out duties not formally part of ones job, and endorsing and supporting organisational objectives.
Once what construes 'job performance' has been adequately defined, measurement of said performance needs to occur. Human judgement is subject to many errors and supervisors are often forced to make some subjective judgements. There are a wide range of problems associated with these judgements. Subjectivity opens the door to favouritism, where evaluators act on personal preferences to favour some employees over others and use their power to reward preferred subordinates beyond their true value. Prendergast Topel (1996) found that degree of favouritism was dependent on the incentives offered to the worker, with a stronger correlation between incentive pay and the supervisors' report decreasing reporting accuracy. The effects of favouritism are a deemphasising of incentive pay and bureaucratic rules in pay and promotion decisions where little weight is put on supervisor appraisals and other subjective opinions of performance and excess weight placed on noncorruptible measures such as seniority.
Rating error, accuracy and bias studies have identified common errors of judgement among ratee's. These include grouping everyone's ratings unreasonably toward the high end of the scale, allowing the 'halo effect' -a favourable judgement on one quality to overshadow other problem areas, and allowing personal bias to affect the evaluation (Rosen 2002). The deliberate inflation of performance appraisal ratings in an attempt to distribute merit pay will decrease the chances of employees with real training needs being identified and increase the chances of overrated employees being promoted beyond their capabilities (Milkovich Wigdor 2001).
Prendergast (2003) "illustrates an incentive for workers to conform to the opinion of their supervisors when firms use subjective performance evaluation" and "that the incentive to conform implies inefficiencies and is likely to lead to more centralised decision making." When the point of reference for adequate performance is the manager's opinion, workers benefit from second guessing the opinion of the manager. An extreme example of an employee trying to manipulate the system in his/her favour is the 'organisational charlatan;' an individual seeking to improve their perceived performance at the expense of their actual performance (Parnell 2001).
A number of recent studies have examined the relative value and interchangeability of different types of performance measures. The most relevant of these assess the relationships between relatively objective and subjective measures of employee performance. Objective criteria, such as production and quality indices, are based on their relevance to desired organisational outcomes (eg productivity, profitability, sales etc.) and measure results, not behaviours. In contrast subjective performance ratings must be based on systematic job analysis, and should be behaviourally based in order to meet technical standards and legal precedents (Hoffman 2001). Meta-analytic investigations of the relationships between relatively subjective and objective measures of employee performance have found only modest differences between the average validities, suggesting the two measures are significantly and moderately related but not totally substitutable (Bommer 1995).
Modern solutions to the subjective nature of performance appraisals take a performance-based approach to the issue. It has been recognised that performance appraisal systems should emphasise elements that one can most 'objectively' measure such as work quality and quantity, accomplishment of goals and objectives and completion of key accountabilities. Subjective factors such as worker personalities and character aren't generally incorporated. A common concern is that management often opts for the quick and easy approach to appraisals and attempts to use the same form for all positions with general and vague criteria.
Markowich (1994) suggests that the evaluation process consists of three phases; identify essential job elements, have a reliable and credible rating system and have a meaningful discussion of the evaluation. This is in the same vein as Escover (2003) who also suggests performance appraisals be conducted as a joint planning activity between management and the employee. The first step involves clarifying work expectations and defining what needs to be done. Quantifiable objectives must be developed so they can be measured and thus an indication of progress or regress made. By definition, if objectives can't be measured they must be rewritten, else their necessity questioned. Weighting of the objectives then occurs by assigning a relative importance to each of the job elements. Various distinct and measurable levels of performance associated with each job objective are established. When the time for evaluation comes around the employee conducts a self-evaluation prior to the evaluation meeting and the discrepancies form the basis of discussion at the meeting.
Research into the relative fairness of performance appraisals and rating systems using concepts of due process and procedural justice, (Folger 2002) has resulted in the development of three characteristics of a due-process appraisal system. These incorporate giving 'adequate notice,' employing organisations should publish, distribute, and explain performance evaluation standards and processes, a 'fair hearing,' a formal review meeting in which an employee is informed of his/her evaluation and 'judgement based on evidence,' raters should apply performance standards consistently across employees, without distortion by external pressures, corruption or personal biases.
Despite most performance appraisals being subjective there is increasing optimism regarding the use of supervisory ratings and other 'subjective' appraisal instruments and formats and it is generally recognised that such systems do not automatically translate into rater error or bias and often provide cost effective solutions to the measurement of performance. In light of studies suggesting the relative validity of subjective measures of performance perhaps the one area that organisations could act to improve is the training of managers and supervisors in how to effectively conduct such appraisals (Hubbart 1995).
References
Arvey, RD Murphy, KR 1998, 'Performance evaluation in work settings', Annual Review of Psychology, vol. 49, pp. 141-169.
Bommer, WH, Johnson, JL, Rich, GA Mackenzie, SB 1995, 'The interchangeability of objective and subjective measures of employee performance: a meta-analysis', Personnel Psychology., vol. 48, No. 3, pp.587-605.
Borman, WC Motowidlo, SJ 2003, 'Expanding the criterion domain to include elements of contextual performance', Personnel Selection in Organisations, Jossey-Bass, San Francisco
Escover, JL 2003, Management Decision, vol. 31, issue. 7, pp. 28-31.
Folger, R, Konovsky, MA Cropanzano, R 2002, 'A due process metaphor for performance appraisal', In Research in Organisational Behaviour, vol. 14, pp. 129-177.
Hoffman, CC, Nathan, BR Holden, LM 2001, 'A comparison of validation criteria: Objective versus subjective performance measures and self- versus supervisor ratings', Personnel Pyschology, Vol. 44, no. 3 (Autumn), pp.601-617.
Hubbartt, WS 1995, 'Bring performance training to life', HR Magazine, vol. 40, issue. 5, pp. 168-170.
Ilgen, DR Hollenbeck, JR 2001, 'The structure of work: job design and roles', Organisational Behaviour Human Decision Process, vol. 54, No: 3, pp. 321-368.
Markowich, MM 1994, 'Response: We can make performance appraisals work', Compensation and Benefits Review, vol. 26, issue. 3, pp. 25-29.
Milkovich, GT Wigdor, AK 2001, Pay for Performance: Evaluating Performance Appraisal and Merit Pay, National Academic Press, Washington.
Parnell, JA Singer, MG 2001, 'The organizational charlatan scale: Developing an instrument to measure false performance', Journal of Management Development, vol. 20, no. 5, pp. 441-455.
Pratt, HJ 2001, 'Principles of effective performance management', ARMA Records Management Quarterly, vol. 25, issue. 1, pp. 28-32.
Prendergast, C 2003, 'A theory of 'yes men'', The American Economic Review, vol. 83, issue. 4, pp. 757-770.
Prendergast, C Topel, RH 1996, 'Favoritism in organisations', The Journal of Political Economy, vol. 104, no. 5, pp. 958-978.
Rosen, DI 2002, 'Appraisals can make - or break - your court case', Personnel Journal, vol. 71, issue. 11, pp. 113-117.
Thorndike, RL 2000, Personnel selection: Test and measurement techniques, Wiley, New York.