Self-rated performance and pay satisfaction, merit increase satisfaction, and instrumentality beliefs in a merit pay environment

The relationship of self-rated performance to pay level satisfac tion, merit increase satisfaction, and instrumentality beliefs in a merit pay envi ronment is investigated. Self-rated performance exhibited a significant negative relationship with all three variables beyond that accounted for by control vari ables. Implications and suggestions for future research are discussed. Merit pay programs typically include a formal appraisal system which yields a performance appraisal score and some mechanism for allocating pay increases based on the appraisal. As a consequence, the accurate assessment of performance is a critical aspect of any merit pay program. Some performance appraisal systems also utilize self-ratings of performance in conjunction with supervisor ratings to determine per formance appraisals. Whether part of a formal appraisal system or in formally, there is little doubt that individuals form opinions about their own performance and that these opinions may differ from their super visor's evaluation of their performance. Indeed, there is evidence to sug gest that employees tend to rate their own performance higher than their supervisors rate it (e.g., Meyer, 1975). It is this tendency for indi viduals to rate their own performance highly that serves as one basis for criticism of merit pay programs. Meyer (1975) argues that most individuals tend to rate themselves Address correspondence to Michael J. Vest, Department of Management, The Univer sity of Southern Mississippi, Hattiesburg, Mississippi 39406. 171 ? 1994 Human Sciences Press, Inc. 172 JOURNAL OF BUSINESS AND PSYCHOLOGY as above average whether they are or not. When an individual employee receives a smaller than expected pay increase based on a supervisor evaluation that is lower than their own assessment of their perfor mance, Meyer (1975) suggests that individuals will react to this threat to their self-esteem by downgrading the importance of the activity or disparaging the source of the criticism, both of which are counterproduc tive to the organization. We suggest here that this tendency for individ uals to rate their own performance highly presents problems for merit pay programs beyond those suggested by Meyer (1975). Lawler (1971) hypothesized that personal input factors such as per ceived job performance (self-rated performance) will affect satisfaction with pay. Individuals who consider their personal inputs (performance) to be high will believe they deserve higher levels of pay, and they will be satisfied with their pay only if the pay they receive is greater than or equal to what they believe they deserve. As suggested by Lawler (1971), if level of pay is held constant, self-rated performance will be negatively related to pay level satisfaction. A similar argument can be made with respect to the relationship of self-rated performance to both instrumen tality beliefe merit increase satisfaction, both variables being important to merit pay programs. Employees must believe pay is tied to perfor mance (instrumentality beliefs) if merit pay is to be successful (e.g., Hills, 1979) and both pay level satisfaction and merit increase satisfac tion have been found to be related to instrumentality beliefs (Heneinan, Greenberger, & Strasser, 1988). As suggested by Lawler's (1971) pay satisfaction model, employees will believe they deserve larger pay in creases if they perceive their performance to be high, and will be satis fied with their pay increase and/or believe pay is tied to performance only if the increase received is greater than or equal to what they be lieve they deserve. Thus, with pay increase held constant, self-rated per formance should be negatively related to merit increase satisfaction and instrumentality beliefs. There is limited and mixed empirical support for the proposed rela tionship of self-rated performance to pay satisfaction, merit satisfaction, and instrumentality beliefs. Both Dreher (1981) and Motowidlo (1982) found a negative relationship between self-rated performance and pay satisfaction while controlling for pay level and demographic characteris tics. Dyer and Theriault (1976), in a study of three groups of United States and Canadian managers, found a negative relationship between pay satisfaction and self-rated performance in the United States and French Canadian manager groups, but not in the English Canadian manager group. While not addressed in the research, it is possible that differences in demographic characteristics may have affected study re sults as the English Canadian group was younger, had slightly less se M. J. VEST, K. D. SCOTT, AND S. E. MARKHAM 173 niority, and was lower paid than the other groups. Failure to control for pay increase or demographic variables may also have contributed to study results. Ash, Lee, and Dreher (1985) found no significant relationship be tween perceived job performance and either pay satisfaction or merit increase satisfaction in a study of part-time graduate students. It is pos sible that using part-time graduate students may have affected study results. For example, respondents may willingly be underemployed with respect to skills and pay because the job allows them to go to school. As a consequence, these individuals may be more realistic with respect to the value of their personal inputs. While pay level was held constant, pay increases and demographic characteristics were not controlled for in their analyses. Goodman and Moore (1976), in a study of blue-collar and manage rial employees, found self-rated performance to be a marginal negative predictor of instrumentality beliefs for blue-collar workers but not for managers. One possible explanation for these findings is that neither pay level nor pay increase were controlled for in this study. While infor mation on salaries were not provided, it is possible that salaries in the blue-collar group may vary less than salaries in the managerial group. Thus, salary may to some degree have been controlled for in the blue collar group. This might account for the marginal negative relationship in the blue-collar group. The literature review indicates limited and often inconsistent find ings with respect to the relationship of self-rated performance to pay level satisfaction, merit increase satisfaction, and instrumentality be liefs. In addition to the sample differences and failures to control for pay level, pay increase, or demographic variables outlined above, findings may also be due in part to differences in measures. For example, while single-item self-ratings were utilized in a majority of studies (Ash, Lee, & Dreher, 1985; Dreher, 1981; Goodman & Moore, 1976), Motowidlo (1982) utilized a 47 item scale which addressed multiple aspects of per formance rather than an overall perception of performance. Similarly, different measures were utilized for pay level satisfaction, e.g. Ash, Lee, and Dreher (1985) utilized both the Job Descriptive Index and the Pay Satisfaction Questionnaire while Motowidlo (1982) used a seven-item scale constructed for the study. The purpose of this research is to investigate the relationship of self-rated performance to pay satisfaction, merit increase satisfaction, and instrumentality beliefs in a merit pay environment. To the extent there is a negative relationship and individuals rate their performance highly, the viability of merit pay programs is suspect. Many employees are likely to be dissatisfied with both pay and merit increase and have 174 JOURNAL OF BUSINESS AND PSYCHOLOGY lower instrumentality beliefs because they rate their performance highly. As a consequence, merit pay may not be a motivator for a large number of employees. The discussion above argues for the importance of controlling for outcome variables such as pay level and merit pay increase when inves tigating the relationship of personal inputs (self-rated performance) to pay-related attitudes. For this reason, pay level, performance appraisal, and merit pay increase are utilized as control variables. Individual dif ference variables such as gender, level of education, and tenure have been found to be related, either directly or indirectly, to either pay satis faction, merit satisfaction, or instrumentality beliefs. For example, gen der has been found to be related to pay satisfaction (Dreher, 1981) and education (Goodman & Moore, 1976) has been found to be related to instrumentality beliefs. As a consequence, gender, level of education, and tenure are also used as control variables due to their potential in fluence on pay related attitudes. Merit pay increases in the research organization were based on both performance appraisal score and posi tion in pay range (via guidechart). As a result, position in pay range is also used as a control variable. Based on the above discussion, it is hy pothesized that self-rated performance will exhibit a significant unique negative relationship with pay level satisfaction (Hypothesis 1), merit increase satisfaction (Hypothesis 2), and instrumentality beliefs (Hy pothesis 3) beyond that accounted for by control variables.