“What’s the use of killing yourself? You still get the same rating as everyone else, and you still get the same 5% increase. It’s demoralizing and demotivating” (Murphy, Kevin J, 1990, p. 4). Does this mindset encourage greater productivity & organization excellence? How much do you feel you’re worth to your employer? Everyone wants to be paid according to how valuable they see themselves. The real question is, how valuable does your employer or co-workers feel you are to the team. Should excellent performance be compensated for? Should unacceptable performance go without reprimand or remedial training efforts? This human resource development solution opportunity focuses on performance appraisals for, a well-known pharmaceutical company (Merck ; Co., Inc) that is faced with the dilemma of revising or overhauling their performance appraisal and salary administration program to combat the demotivating mindset that has become the consensus among its exempt salaried employees. Should the performance appraisal and salary administration system be revised? Where do the problems exist and what changes can be made to correct this mentality plaguing the organization?
In March 1985, a newly appoint CEO Dr. Roy Vagelos formed the Employee Relations Review Committee charged with evaluating the personnel policies and practices that contributed to a recent lag in productivity, despite the obvious disappointing new products, inflation, and changes in foreign currency. Regarding personnel practices, “compensation for exempt salaried employees at Merck has traditionally ranked among the top 25% of large U.S. companies. Merck’s progressive personnel policies and aggressive pay practices have contributed to high levels of employee loyalty as characterized by historically low turnover rates” (Murphy, Kevin J, 1990, p. 2). Merck’s existing performance appraisal and salary administration program was a revision of the original program, introduced in 1978. Under this plan, the supervisors rated employees on a one to five scale under multiple rating categories. The design of the current plan leaves room for bias errors such as the halo effect. A positive or negative halo affect occurs when a rater generalizes an employee’s good or bad behavior on one aspect of the job to all aspects of the job. In Merck’s case, the error of central tendency is committed when raters “play it safe” by keeping employees in the middle of the scale to avoid a low or high rating (Bacal & Associates, 1998). The outcome of this affect harms both the outstanding and poor performer. Merck seemed to have policies in place, but they were viewed as less than effective, due to improper application. In this case, the current appraisal and salary administration system was not being applied correctly by the leadership. To go from the cutting edge, top grossing, powerhouse that Merck & Co. was, to declining sales and lackluster performance, forming the Employee Relations Review Committee was definitely needed to diagnose the symptoms of a broken system.
It is better to analyze the specific strengths and weaknesses of the current systems prior to discussing alternatives for alleviating the weaknesses. In other words, identify what you may want to keep and then discuss what is not working with regard to appraisal or compensation systems.
To begin, we must note that there are several plausible alternatives to evoke change within the current system, but are they all suitable to the current situation? Performance appraisal methods fall into four broad categories: trait systems, comparison systems, behavioral systems, goal-oriented systems. The traits approach “ask raters to evaluate each employee’s traits or characteristics (e.g., quality of work, quantity of work, appearance, dependability, cooperation, initiative, judgement, leadership responsibility, decision-making ability, or creativity)” (Murphy, Kevin J, 1990, p. 60). The main disadvantage with the trait approach is that it is highly subjective based on the supervisor’s perception. In the comparison approach, supervisor’s build a hierarchy by ranking employees in comparison using overall performance or various traits. Forced distribution performance appraisals are alternative to the comparison approach that assigns employees to groups that represent the entire range of performance. This approach can be useful in Merck’s case, because it minimizes the tendency for the supervisors to rate most employees as excellent performers. A problematic disadvantage of force distribution is that the Management-Employee relationship ultimately suffers because the workers feel that the ratings are unreal in comparison to their individual performance. Congruent with the trait system, the comparison approach also has limitations. The comparison approach tends to encourage subjective judgement, which increases room for error by the rater due to errors or biases. Merck is currently suffering, because they have “hypothetically speaking” put the entire deck in the hands of the supervisors.
In contrast to comparison & trait methods, behavioral systems rate employee’s objective job behaviors displayed daily. When developed & applied correctly, the behavioral approach is virtually free of rater errors and biases. The main behavioral systems are the critical incident technique (CIT) & Behaviorally anchored rating scales BARS. CIT require the supervisor & the employee to differentiate incidents of successful performance from unsuccessful ones. CIT tends be useful because of extensive documentation that identifies the employee’s performance, but this can be burdensome & time consuming on the supervisor. BARS are based on CIT, and developed on the premise, but with one exception. A critical incident is written that incumbent completed the task in a timely manner, but the BARS format is written that the incumbent is expected to complete the task in a timely manner (Murphy, Kevin J, 1990). Among the various appraisal techniques, BARS is most defensible in court, because it’s based off actual recorded behavior. Similarly, BARS is difficult to maintain, because of the constant volume of data required to make the system effective. Jobs change often, the documentation must be update with each change.
Of the four appraisal types, goal-oriented systems seem to be the most plausible. “Management by objective (MBO) could be the most effective performance appraisal technique because supervisors and employees determine objectives for employees to meet during the rating period and employees appraise how well they have achieved their objectives” (Murphy, Kevin J, 1990, p. 64). In Merck’s case, management by objective can prove to be very effective because it involves the input of the employee and the supervisor. On the downside, management by objective is also time consuming, but on the upside, it promotes effective communication between the employee and their supervisor. In Merck’s current situation, transparency is definitely an aspect needed to change the culture among the management staff & employees.
Next, just to paint a mental picture of a relatable situation, we can use the revamping of the US Army’s evaluation system as an example. Some of us are currently serving on active duty in the US Army and in senior level leadership positions. We required to receive annual performance evaluations in-line with our peers that have achieved a certain rank in the Army. Although we don’t use the merit pay strategy per se, these same evaluations have a direct impact on whether we are selected for promotions, which in turn, means we get a pay increase. Promotions can be very competitive, however, it is well known that many supervisors in the Army were creating the exact same issues as the supervisors at Merck & Co. Army supervisors were giving a lot of their average and sometimes, underperforming subordinates an “among the best” rating; the equivalent of a 5 rating at Merck & Co. simply to save face, which caused the Army to promote many Soldiers that simply weren’t capable of excelling at the next rank.
So, two years ago, the Army initiated a major overhaul of our evaluation system, in order to force supervisors to give a more accurate depiction of which Soldiers are truly deserving of promotion. Army leaders forced this change by limiting those that can be selected for a top rating to 24% of the personnel that they evaluate. So, if a supervisor is evaluating 5 people, that supervisor can only give the top rating to 1 person (5 x 0.24 = 1.2). It immediately created more competition amongst the Soldiers, however, it also created a stronger need for us to operate on a higher level. It has also forced supervisors in the Army to conduct real, honest evaluations of performance and finally give credit where credit is due or explain where the Soldier is lacking. Isn’t this more a forced-distribution system than a management by objectives approach? Or, are you not linking this section to your comments in the preceding section about the value of MBO? Needs clarification!
Consequently, we feel that Merck should attempt a similar strategy. We’re not recommending a complete overhaul of the system, more so an aggressive revision of the policies that are in place. The base system in place is an adequate starting point to implement an effective merit pay program, but there needs to be checks & balances in place to enforce the plan. Following the guidelines set forth through the goal-oriented system, we would propose the management by objectives (MBO) approach, because it limits the subjectivity involved in the appraisal process by getting the employee involved in his or her own appraisal. Managers in organizations should recognize the importance of employee involvement in setting performance standards at the beginning of the appraisal period and providing performance feedback during the appraisal period (Inderrieden, Keaveny, & Allen, 1988). In addition, we would also propose strengthening the link between pay and job performance. In particular, our methods need to ensure that there is a clear differentiation between excellent performance & poor performance. Merit pay practices have been shown to be related to improved merit pay plan effectiveness in clarifying the link between pay and performance for the employee by increasing the frequency of appraisals, establishing developmental action plans, and developing a formal merit pay policy with safeguards to ensure employee perceptions of fairness (Eskew & Heneman, 1996).
Finally, it boils down to forcing the managers to do their jobs by awarding merit pay increases based off performance. In practical terms, managers should award the largest merit pay increases to employees with the best performance, and they should award the smallest increases to employees with the lowest acceptable performance. The difference between these merit increases should be approximately equal to the differences in performance (Martocchio, 2017). For example, assuming the median salary for Merck is $91,498.50, if Merck were to revise their appraisal and performance system that pays $18,300,000 for base salary for 200 employees exempt from FLSA provisions and allotted themselves a $1,281,000 budget for a 7% merit increase. If all Merck employees that receive a 4+ or 5 rating, they should receive a 7% salary increase, all 4- or 4 employees receive a 4% increase, and 3+ or lower employees won’t receive an increase at all, because that means that they are simply executing the job they are paid to do. A 3+, 3 or 3- wouldn’t be necessarily a bad rating, it just means that they’re average performers. Any rating below that should be grounds for probation and/or termination. We feel that higher quartile employees are paid higher wages because that means that they add that much more value to the organization. Those are the people that any organization truly needs in order to operate more efficiently. So, we feel that a quartile 4 level position should be granted the same opportunities to earn a 7% increase as well if they earn it, however, it will need to be limited to a stricter rating cap, i.e., 10%, to avoid more than one or two high salary employees getting the top ratings at one time. Everyone could be eligible for the merit increase, because extra funds won’t be spent on those that are rated as average employees. So, budgeting would go as follows, assuming the factors below–
Performance distribution ratings:
Excellent Above Average Average Poor
10% 25% 60% 5%
Percentage of employees whose pay falls into each pay quartile:
Quartile 4 Quartile 3 Quartile 2 Quartile 1
10% 25% 35% 30%
Combine both sets of information to determine the percentage of employees who fall into each cell:
Excellent Above Average Average Poor
10 x 10 = 1 25 x 10 = 2.5 60 x 10 = 6.2 5 x 10 = .5
10 x 25 = 2.5 25 x 25 = 6.25 60 x 25 = 15 5 x 25 = 1.25
10 x 35 = 3.5 25 x 35 = 8.75 60 x 35 = 21 5 x 35 = 1.75
10 x 30 = 3 25 x 30 = 7.5 60 x 30 = 18 5 x 30 = 1.5
Calculate the expected number of employees who fall into each cell:
Excellent Above Average Average Poor
200 x 1% = 2 200 x 2.5% = 5 200 x 6.25% = 12.5 200 x .5% = 1
200 x 2.5% = 5 200 x 6.25% = 12.5 200 x 15% = 30 200 x 1.25% = 2.5
200 x 3.5% = 7 200 x 8.75% = 17.5 200 x 21% = 42 200 x 1.75% = 3.5
200 x 3% = 6 200 x 7.5% = 15 200 x 18% = 36 200 x 1.5% = 3
Expected number of employees in cell × Desired pay increase for cell (%) × Current median pay level for the current quartile:
Quartile 4 Quartile 3 Quartile 2 Quartile 1
$91, 498 $76, 498 $61, 498 $46, 498
Excellent Above Average Average Poor
2 x 7% x $91,498 = $12,809 5 x 4% x $91,498 = $18,299 12.5 x 0% x $91,498 = $0 1 x 0% x $91,498 = $0
5 x 7% x $76,498 = $26,774 12.5 x 4% x $76,498 = $38, 249 30 x 0% x $76,498 = $0 2.5 x 0% x $76,498 = $0
7 x 7% x $61,498 = $30,134 17.5 x 4% x $61,498 = $43,048 42 x 0% x $61,498 = $0 3.5 x 0% x $61,498 = $0
6 x 7% x $46,498 = $19,529 15 x 4% x $46,498 = $27,898 36 x 0% x $46,498 = $0 3 x 0% x $46,498 = $0
The sum of expected merit pay increases equals $216,740.
I like your approach overall, but it still doesn’t tell me how to address those employees already at or near the maximum of the pay grade. How much over the market rate are you willing to pay people for doing a similar job? This is why, as I stated in my announcement comments that many organizations have gone to an across the board COL increase annually for all employees and then distributed bonuses to exceptional performers using the appraisal system you have proposed. Since bonuses are one-time awards, they are not added to salary so individuals approach the pay grade maximum much more slowly.
Overall, while never achieving perfection, pay for performance systems can serve their purpose if used in conjunction with a well monitored evaluation system. Supervisors will need to be adequately trained on how to implement the revised appraisal system and understand that everyone should be evaluated on their ability to perform their job efficiently and effectively, not their personality traits or popularity around the workplace. Leaders/supervisors get paid a higher salary specifically to lead. They are the first level of checks and balances that need to occur to ensure its ultimate success. However, just in case some supervisors are resistant to change, the Employee Relations Review Committee needs to stay intact to conduct random assessments of policies and employee morale to foster a continuous environment of growth.