Employers Need to Invest to Strengthen Performance Management

How We Got to Where We Are

For those who were not involved in HR before 1990, it was a very different era. Performance systems used in the 1980s looked very similar to those that were developed in the years after World War II; not much had changed. Actually the ideas go back even earlier. The first PM policies originated in the early factories and were intended primarily to confirm an employee met expectations. That philosophy prevailed through the 1980s. Managers and supervisors saw PM as a practice imposed and required by HR. The idea was never truly accepted as useful.

Dick Grote, a prominent advocate for PM, is also quoted in the article. He argues—and I agree—that it is a necessary mechanism for personnel decisions—promotions, terminations, pay increases and so on. The caveat is that employers should be able to defend the ratings behind those decisions. That is often questionable.

My guess is that the defense of ratings will become increasingly important. I am sure I am not alone in anticipating the issue of pay discrimination and therefore transparency will be important in Hillary Clinton’s presidential campaign and, if she is elected, in her presidency. If that is accurate, there will be a heightened need to defend pay differentials and the basis for making increase decisions.

For years HR heads reacted to dissatisfaction with PM by adopting a new form. That, however, was only a short-term fix. The dissatisfaction reoccurred and the cycle repeated. Research by psychologists tried to develop defensible rating methods but their studies ended long ago. Goal setting is clearly important but it is not applicable to all jobs and can be gamed. Goal setting can never be managed equitably across an organization. Computers facilitate collecting and assembling information but clearly technology has not solved the core problems. My sense is employers now need to truly address the weaknesses in many of the off-the-shelf answers.

This would not be an issue if there were objective performance measures available for all jobs. Increasingly employers are developing metrics but objective measures at the job level are typically unavailable. The use of performance goals is helpful if they include agreed upon measures. In the absence of quantitative measures, performance criteria should be verifiable. Mutually accepted performance measurement scales are a possibility. But for the majority of jobs, performance assessments are subjective and that is unlikely to change.

It is complicated by the changing nature of jobs and organizations. There are many employees who perform their job in locations away from their supervisor. Moreover, with knowledge jobs, it is not possible to observe an employee as supervisors did in a traditional manual work setting. As I write this, someone asked the question on a website, “What would it be like to manage remotely the workforce just from your computer?” No single PM system can satisfactorily “fit” the various work settings in a typical organization.

Research by Gallup and others confirm the obvious. Employees want to be treated fairly; they want to know what is expected and what they can expect, they need feedback to improve their performance and they want to be recognized for their contribution. Ideally they want their performance to lead to career progress and rewards. The thread that ties all of that together is the central importance of communications between managers and their people.

Managers and human nature contribute to a core problem—the tendency to give employees an inflated rating. That is a rating that is not warranted by their performance. Government is no doubt the worst; it is not uncommon to see 90% of the ratings at the highest levels. Human nature also prevents managers from rating employees as unsatisfactory. They are simply uncomfortable giving a subordinate bad news.

What is sometimes not recognized is that someone has to make those people decisions and it should be an individual who is accountable for performance. That is not HR’s responsibility, and it is not the PM system; it is always a line manager at some level. We should have recognized years ago that managers are at the heart of the problem.

This is not a discussion of how to plan and management performance systems. A useful discussion is provided in the SHRM booklet: “Performance Management: A Roadmap for Developing, Implementing and Evaluating Performance Management Systems.” The author, Elaine Pulakos, PhD, has published frequently on the subject. (She and I are both Penn State grads.) There are many books on the subject.

The Purpose of Performance Management

Part of the problem is that PM serves several purposes. Traditionally, it has been the source of information used by decision makers for pay increases, promotions, terminations and other job- or career-related decisions. It is also important for identifying high-potential employees. Those decisions are covered by employment law and there is a need to defend the decisions as nondiscriminatory.

Ratings are used by an employee’s immediate manager as well as by other managers and committees for staffing and organizational planning. Employers have special policies to recognize their best performers.

Now PM is defined as a process that starts with performance planning at the beginning of the year and ends with a year-end evaluation. The planning is basic to effective management. Throughout the year it is useful to a job incumbent to have feedback and coaching to improve performance. If the job involves performance goals, measurement systems can also provide periodic feedback as the year unfolds. Coaching is a new and still not widely acknowledged role for managers and one that is less important in the minds of organizational leaders.

There is evidence that younger workers are more interested in feedback and career progress. Older workers would no doubt have benefitted from regular feedback but it was not expected from managers early in their careers. The argument is that the role of being an effective performance manager/coach makes it awkward to make rating decisions. Managers are uncomfortable providing honest feedback that is negative despite the fact that it is necessary to help an employee improve.

Apparently (perhaps there is research that I have not read) the presumed conflict in the two roles is exacerbated if performance ratings early in a career are inflated—as they often are. That is to say, it makes it more difficult to justify providing negative feedback later in an employee’s career.

But there is no inherent reason why a well-intentioned and adequately trained manager cannot serve adequately in both roles.

The Analogy of Testing in School and Rating Reliability

Occasionally articles on PM use the analogy of testing children in school. I will use that here to discuss aspects of the solution to the PM problem.

In school each teacher has a somewhat unique style and orientation to the subject. Two teachers teaching the same subject will cover slightly different material and discuss the material from a different perspective. The effectiveness of teachers can vary widely. Children have different learning patterns and perceive their teachers differently. They might see a teacher as a “tough grader” or another as “an easy A.”

The educational community recognizes the importance of teaching skills—pedagogy—and has course requirements along with student teaching to ensure each new teacher has prior training. The obviously ill-suited future teachers are counseled to find a different career before they enter the profession. Even with that, parents obsess sometimes over the effectiveness of their child’s teachers. In contrast, supervisors often have minimal training and little if any oversight.

Teachers frequently have a “favorite” student and of course children know it is advantageous to “brown nose.’ It is no doubt human nature to react to some individuals more favorably than others. Bias and discrimination are known to affect PM ratings.

Teachers also have the advantage that they planned for a career in teaching. For many it was a career goal from the time they were in school. Managers and supervisors, in contrast, sometimes take the job only to make more money or are given the job through seniority. Too few employers select new supervisors based on a valid assessment of their potential to make the career change successfully.

The effectiveness of managers and supervisors varies as much as individuals in teaching. Significantly, it is the differences in the way they handle their responsibility for managing employee performance that contribute to the PM problem. They vary in (1) their commitment to using the PM system as HR planned it, (2) their personal “grading” philosophy, (3) their commitment and skill in observing and assessing employee performance and (4) in their relationships with employees. The bottom line is that—it is unrealistic to expect them to be consistent in the ratings. Two managers can observe an employee perform his or her job and evaluate the performance differently. That is a problem.

Consistency is a fundamental requirement to fair and defensible PM ratings. In school a teacher might have 20 to 30 students so at least they can expect a degree of consistency in the teaching and in final grades within the class. Across an organization one manager might have to rate 10 employees while others might have only one or two. Even employees with the same job title often have differences in job duties. To extend the analogy, each manager is essentially the same as a separate classroom.

It is for that reason that I am convinced the use of calibration committees is essential. These are committees of peer or next-level managers who assess “draft” ratings to insure fairness and consistency. Each manager is expected to explain and justify his/her draft ratings in a meeting of the committee.

Committees have been used for over a decade. There are many variations in their composition and in way they are used. Research confirms they work—they significantly improve the credibility and the validity of ratings. Employees are more satisfied they have been treated fairly. The practice exposes a broader group of managers to performance issues across an organization, enables them to become familiar with the high potential employees and reinforces the recognition value of the ratings.

The new book by Laszlo Bock, Google’s head of People Operations—Work Rules!—devotes most of a chapter to their decision to use calibration committees.

The School Analogy and the Use of Generic Performance Criteria

The school analogy also helps highlight another fundamental problem with the use of PM. The early PM systems had a limited purpose—to confirm employees met what in reality are minimal expectations, did what they were told and stayed out of trouble. Supervisors were required only to check off a few boxes; it was perfunctory and seen strictly as “more HR paperwork.”

That mindset unfortunately has continued in many work settings.

The early appraisal forms focused on generic criteria like “diligence,” “honesty” and “cooperative.” The individuals who plan PM systems have introduced new ideas related to the “best” criteria over time but there was no gain in user satisfaction. In the late 1980s, the idea of competencies was introduced for selection. I first saw competencies used in PM in 1990 for nurses at the federal medical center, National Institutes of Health. Competencies are now widely used, but because there is no standard definition it is not clear how widely they are used or how effective the practice may be.

But the practice today too often continues to reflect the earliest thinking in one regard—employers rely on the same performance criteria for obviously different job families. If the only purpose of ratings is to confirm an employee met minimal expectations, that really does not matter. It gives the rating process a patina of fairness, everybody is treated the same.

However, if the purpose is to give an employee feedback on his/her strengths and weaknesses to help improve performance, then it follows that the feedback and the criteria have to be job specific. Here a sports analogy is useful. To evaluate a quarterback, the criteria should be specific to the position; the same criteria would not be useful in evaluating a tight end or a tackle.

In the same way, the performance criteria relevant to an accountant or an engineer are not the same as those relevant to HR.

In school the same test would never be used, for example, in both math and history.

Using the same logic, the same math test would not be used for students in 7th and 10th grades. In a work setting the performance expectations and the criteria for a senior engineer would not be the same as those for a new graduate. It is also true that in a school with multiple levels of student ability in the same grade, a group of honors students would have a different exam than C students.

To reiterate, the purpose of the evaluation is to document how well an employee performs his or her job. That can only be accomplished if the feedback and the ratings are job specific. Surveys show employees frequently state the process is not useful for improving performance. That in part is attributable to relying on generic performance criteria.

It was in the 1960s that the idea of goal setting was introduced. Now of course goal setting is central to PM for executives, managers and most professionals. Goal setting is by definition job specific. It can be gamed—with experience it is generally possible for individuals to negotiate goals that are easy to beat—and it takes experience and coaching from proven users to develop goal setting skills. It is not perfect but it is credible and proven to contribute to improved performance. (I discussed goal setting in a prior editorial.)

Defining Performance Expectations

A weakness in many PM systems is the implicit, limited goal of meeting expectations. That goes back decades to those early PM approaches. It happens when only one level of performance is defined—“satisfactory.”

My experience has convinced me that employees typically want to be seen as good performers, and to have the information that helps them to be star performers. Many aspire to be rated as stars and will put forth the effort to achieve that rating.

A potential problem is that employees have enjoyed inflated grades in school and parents and others have told them repeatedly that they are above average. Lake Woebegon is real! In part this is fed by the experience many students have in school where grade inflation it seems has become common. (I am convinced it is easier now to earn those As than when I was in school.)

It is also too easy for a manager to rate an employee as a 4 or 5 (on a five-point scale) when expected performance is only defined at the “3” level. The reality is it is so common the problem is usually ignored. It explains why there are so many employees at the 4 or 5 level.

It is also the reason GE years ago adopted a forced distribution policy, where only 20% could be rated as outstanding and 10% had to be rated as unsatisfactory. The idea gained popularity because so many employers have problems with inflated ratings. Fortunately most of “rank-and-yank” or forced distribution policies have been eliminated. The practice cannot be defended. It causes morale problems.

A useful strategy is defining performance at three levels—unsatisfactory, satisfactory and outstanding or role model. The words can be replaced by others. I prefer using “role model” as it has a connotation of truly being among the best at something. They can be used with a three-level rating scale (as used with forced distribution policies) or with five-level scales. At the lowest level, the bottom level makes it far easier to justify terminating a poor performer. At the highest level, the rating has recognition value, reduces inflated ratings since it is hard to justify and provides a goal for those who aspire to be the best. Generally high performers in a job family can define all three levels for relevant criteria in a meeting or two.

The Use of Multiple Rating Sources

Multiple rating sources are no doubt necessary in virtually every organization. It was only in a traditional work setting with a supervisor in close proximity to employees at work that a single rating source ever made sense. But today a typical manager has few opportunities to observe employees at their jobs.

Performance measurement systems can enable managers to monitor results—the data—but that does not provide an understanding of how an employee interacts with customers, coworkers and so on. There are multiple vendors with software for collecting and summarizing from people who are in positions to comment on an employee’s performance.

The recommendation is to determine for a position who would be able to provide useful feedback. Furthermore, a short list of specific questions should be developed for each group invited to comment. That is to say, a customer should be invited to respond to different questions than a coworker. Both the manager and the employee should see summary response data. The manager remains responsible for the evaluation decisions.

The Importance of Gaining Buy-In

Too often when PM is seen as an HR responsibility, no one—line managers or their people—sees the process as valuable. Surveys frequently confirm it is barely tolerated.

Employees want to know they will be treated fairly. They need to see the process as useful and ideally valuable to their career plans. They are far more likely to feel that way if they or their coworkers are involved in the planning. Job incumbents know their jobs and the performance criteria better than anyone. With guidance, a small group of experienced, high performers can identify and define the criteria in two or three meetings. Their work product will more than likely be accepted by their peers.

Equally important, managers need to understand PM is an important part of their jobs. Top management needs to make it a priority. Managers need to be involved in the planning. They also need adequate training to assure they have the feedback skills. Their effectiveness should be considered in their evaluation and in determining their rewards. The best should benefit; those who are ineffective need to be moved to a nonsupervisory role. That will send a powerful message.

The Bottom Line: Managers Have to Be Accountable for Performance

Performance management is not an HR problem; it is a management problem. HR may need to convince organizational leaders that PM should be a priority and to provide leadership in evaluating current practices and in guiding the planning for a new approach. Since there will be a need to involve technology, HR should be involved with the selection of a contractor. From that point HR will no doubt provide training for managers, handle employee communications, keep the records and possibly provide coaching for managers—but managers will need to make the system a success. HR cannot do that.

Their handling of their PM responsibility can be assessed with two approaches—one is the monitoring of performance results and the second is by surveying employees.

HR should assume responsibility for analyzing rating patterns for possible bias and discrimination. That of course should be an issue considered by the committees.

The best managers should be rewarded. Those who prove to be ineffective should be moved back to nonmanagerial roles. The cost is too high to allow them to remain as managers.

Recent research by Wharton’s Adam Grant confirms what we know but may not have focused on. In any large organization some managers inspire considerably higher performance than others. When they move within the organization to a new position, their new work group will in a relatively short period show performance gains. It is the manager and his/her ability to motivate that triggers the improved performance. Grant’s research shows middle managers have more impact on performance than any other factor. Effective performance management is a key. The investment will pay off.