What Can We Do Instead of Appraising People?
Alternatives from a 1987 Paper by Peter Scholtes
Performance evaluation is an exercise in futility. It is an activity that lulls us into the belief that we understand something, when all we have accomplished is to create an oversimplified illusion about something that is very complex. When we act as though our evaluations are accurate, when we reward, punish, promote, commend or retrain people based on our evaluations, we are making adjustments to a system about which we know very little.
- Scholtes, Peter. An Elaboration on Deming’s Teachings on Performance Appraisal. (p. 20)
Deadly Disease #3: APPRAISAL OF PERFORMANCE
The effects of performance appraisal (personal review system, merit rating, evaluation of performance, annual review, system of reward, pay for performance, etc.) are devastating.
“Management by Objective” is a similar evil; “management by fear” would be an even more appropriate title. The results of such practices are as follows:
They nourish short-term performance, rivalry, and politics; the annihilate long-term planning, build fear, demolish teamwork;
They leave people bitter, others despondent and dejected, some even depressed, unfit for work for weeks after receipt of rating, unable to comprehend why they are inferior. They are unfair, as the ascribe to people in a group differences that may well be caused totally by the system within which the group works.
- Dr. Deming as quoted in The Deming Dimension (pp. 50-51)
Differences there will always be between any two people, any two salesmen, etc. The question is, what do the differences mean? Maybe nothing. Some knowledge about variation (statistical theory) is required to answer these questions.
- Deming, Dr. W.E. The New Economics, 3rd Ed. (p. 19)
THE AIM for today’s newsletter is to share with you some excerpts from a 1987 paper by Deming contemporary and collaborator, Peter Scholtes, which he wrote while a consultant for Joiner Associates. Entitled An Elaboration on Deming’s Teachings on Performance Appraisal, it aims to pick up from where Dr. Deming left off and make a forceful case on why appraisals are misguided and what can be done instead. While written over thirty-six years ago, the lessons are just as important and relevant in today’s modern workplaces.
I am grateful and indebted to newsletter subscriber, Victor Vukorpa, for providing me a copy of the paper, which can be downloaded here.
tl;dr: Don’t Appraise People
Scholtes doesn’t equivocate about looking for an alternative practice that simulates appraisals in another form: “Just stop doing them… like drugs [they] are demonstrably the wrong thing to do. They cannot accomplish what they are supposed to do. You do not need a substitute for performance appraisal in order to discontinue it.”
What’s the Problem with Performance Appraisals?
In simpatico with Dr. Deming, Peter Scholtes was staunchly opposed to performance appraisals as a practice principally because they misattributed performance solely to individual without considering the impacts of the system they worked within. In this paper, however, Scholtes takes a deeper dive into the mechanics of appraisal, exposing how and why they fail in their intended purpose.
He begins with a deconstruction of the atypical employee performance process, comprised of three inputs as shown below:
Within each of the inputs are several contributing factors that are predicated upon an assumption or preconceived notion about what contributes to providing an “un-biased and objective” evaluation. As shown below, Scholtes illustrates these using a fishbone or Ishikawa diagram to expand the cause/effect relationships. I’ve added color-coding to conceptually link this to the process diagram above, with orange highlights indicating influences from the host organization’s system:
In turn, each of the influences are themselves based on a flawed set of premises and assumptions that effectively reinforce the notion of sorting out defects from the “talent pool” and either retraining/rehabilitating or rewarding who remains. Scholtes’ theory here is that this is because our perceptions of related traits and characteristics are mismatched against the likely reality in our organization: we “see” wide distributions of traits and characteristics where they are probably narrow, and vice-versa, as shown below—I’ve aggregated these from the paper and highlighted to correspond to the above diagram for clarity:
For example, take the assumption of “Native Ability”, where the theory is that we assume a broad distribution among employees that corresponds to our preconceived ideas between people who “exceed” and “do not meet” expectations in a role: you are either a “rockstar” or a “dullard”, and the job of the appraiser is to sort one from the other for commensurate rewards and promotions or PIPs and dismissals. (NB: Consider what this says about your hiring practices…) The reality, however, is that the difference in talent is much more narrowly distributed, with perceived differences arising from the systems and attendant processes the individual works within.
The pattern in the theory inverts with the “Variation in Methods of Evaluation” distribution, where we presume that the majority of our methods of evaluation are quite uniform and fair regardless of who interprets and uses them, with few outliers that are too lax or harsh. In reality, regardless of who designs the methods, the inherent differences between evaluators, ie. variation, means the results will similarly vary in a wider distribution. Caveat: Some HR policies deliberately “shape” this distribution by limiting the number of “exceeds” that can be awarded, irrespective of what the manager may have observed, as a cost-containment measure.
Combined, Scholtes argues that these mismatches between perception and reality render conventional performance appraisals less an unbiased, objective assessment of merit and more of an arbitrary lottery with inherent biases that can leave employees bewildered, up one year and down the next, despite thinking they’ve been doing their level-best when it was never really about “their” performance in the first place. An additional confounder is the Pygmalion Effect where one’s prior rating tends to inform the next, ie. high ratings beget high ratings, low ratings beget ever more low ratings, which we discussed in our March 28/22 newsletter.
In sum, the practice conflicts with reality, demoralizes both manager and managed alike, and fails to live up to expectations. Why do it?
What Could We Do, Instead?
Scholtes is blunt in his primary advice: “Just stop doing them… like drugs, performance appraisals are demonstrably the wrong thing to do. They cannot accomplish what they are supposed to do… You do not need a substitute for performance appraisal in order to discontinue it.” While this might seem radical, it has been done many times by many different people in many different organizations, with some never instituting the practice in the first place.
However, most organizations have perceived dependencies and needs that could go unmet without an appraisal process; what then? Scholtes lists eight anticipated needs and corresponding remedies at the end of his paper:
Performance Appraisal Provides Feedback to Employee:
Identify the major processes that each employee is involved with;
Identify the major work groups they belong to;
Develop lists of 10-15 major feedback resources for each employee;
For each feedback resource, develop an agenda and method for soliciting that feedback.
Performance Appraisal Provides a Basis for Salary Increases and Bonuses:
Pay a fair, market rate to begin with;
Pay for the accumulation of related skills;
Pay for the accumulation of responsibility;
Reward seniority and the contributions that results in;
Institute profit sharing company-wide.
Performance Appraisal Identifies Candidates for Promotion: As the saying goes, past performance is not an indication of future results, and appraisals are biased.
Design special assignments that contain elements of the next role; supplement with coaching, with the aim for the candidate to succeed;
Develop skill assessment centers to directly observe candidates as they practice the required skills under simulated conditions of the next role; supplement with coaching and help;
Find ways to involve the “customers” of the next role in developing the criteria and selection methods for deciding best-fit;
Develop an organizational culture that is less dependent on promotions as a proxy for leadership; find ways and means to reward without having to climb the ladder.
Performance Appraisal Gives Periodic Direction to an Employee’s Work:
Dedicate time to develop and communicate the organization’s mission and operating philosophy. This is all about AIM and CONSTANCY OF PURPOSE.
Managers should spend extensive amounts of time with their employees planning over successive time horizons: the next project, the next few months, the next year, then three, five, and ten years. The aim is to provide the “how” to the goals.
Communicate with employees every day to get a good sense of what is going on and how tasks are progressing. There shouldn’t be any surprises for manager or employee.
Performance Appraisal Provides Opportunity to Give Recognition, Direction, and Feedback to An Employee Regarding His or Her Work on Selected Responsibilities and Special Projects.
As with prior points, the prescribed remedies are: teamwork, planning, getting feedback from customers and other key resources, communication, communication, and more communication.
Performance Appraisal Identifies an Employee’s Need for Training, Education, and Skill or Career Development
Understand the job to be done and the requirements, then design methods to determine people’s capabilities for the necessary competencies. Enlist expert help to design training to match.
Leverage casual communication with employees to identify and explore educational opportunities.
Make education a corporate obsession rather than a cheap, often cut-out “benefit”.
Performance Appraisal Provides an Equitable, Objective, Defensible Promotion System: It satisfies the requirements of the 1964 Civil Rights Act and Equal Opportunity Commission Guidelines of 1970.
Per above, conventional performance appraisal only provides the appearance being an equitable, objective system: it is neither, and therefore indefensible.
Ask: Are we committed to the spirit and ideals of equal rights and equal opportunity? Are we committed to helping improve the job and career prospects of those in groups who in our society have been victims of discrimination?
Design well-planned initiatives to ensure fairness for everyone; see prior remedies for suggestions.
Performance Appraisal Provides a Channel for Communication that would Probably Not Occur.
Ask: Why has this communication not otherwise occurred? What could we do to provide this sooner when it is most relevant and useful?
Edit: Scholtes also mentions in a subsequent section a practice for “alternatives to effective communication” that Dr. Deming would have disagreed with: “Management by Walking Around”. At the time, Lean was quite new to North American business and there was a good deal of naïve copying of practices such as Gemba walks, and I imagine Deming wasn’t keen on furthering the misconception.
In Out of the Crisis, he wrote:
MBWA (management by walking around, a term that I learned from Lloyd S. Nelson) is hardly ever effective. The reason is that someone in management, walking around, has little idea about what questions to ask, and usually does not pause long enough at any spot to get the right answer.
Reflection Questions
Consider the arguments that Peter Scholtes puts forth in his paper regarding the problems with and remedies for performance appraisals: Do you find merit in his assessment that the practice is mismatched to reality, ie. is it more about keeping up appearances than being effective in its stated aim? Can individual performance be disentangled from the system(s) and processes they work within? Is it possible to design a completely fair appraisal system free from evaluator bias?
Consider the Perception vs. Reality distributions that Scholtes provides to support his theory that we tend to be overly-optimistic or pessimistic in our assessments of the traits and characteristics of people and the organization’s systems. Do you agree? What other distributions might there be in your own organization?
Consider how performance appraisals are conducted in your own organization: Do they match with Scholtes’ assessment, or are they different in substantive ways? How so? What would happen if they were to be eliminated? What needs would have to be fulfilled in other ways? Would any of the remedies Scholtes lists be applicable? What others could you suggest?
I just wanted to say a huge thank you for writing such an insightful piece. Your article on alternatives to traditional performance appraisals really got me thinking. It's refreshing to see someone tackle such a hot topic in a way that's not just informative, but genuinely engaging.