By definition, most people perform their jobs at an average, or somewhere near an average, level. Most, however, do not want to receive average performance evaluations. They want to get "fives" on a five-level scale. Federal managers have tended to oblige this desire, giving most employees top ratings.
Whether in response to such criticism or to avoid federal employees' distaste for low ratings-and the paperwork necessary to justify low ratings-many agencies in the past decade have switched from five-level performance evaluations to simple pass-fail systems. Nearly half of federal employees are now rated pass-fail. Very few fail.
Alas, many say pass-fail systems encourage mediocrity and do not recognize star employees' work over that of their peers. So some agencies have switched back to or are considering switching back to five-level scales, or perhaps to systems with three or four levels.
Federal law requires managers to rate employees on job elements and standards, so some sort of system has to be in place. Several observers point out that managers intuitively group their employees into three categories: stars, who make up a small percentage of employees; good employees, who are the majority of the workforce; and turkeys, or deadwood, who constitute the smallest group. That is why some management consultants see a trend toward three-level systems.
But the obsession with the number of rating levels strikes some federal managers and employees as pointless. "It really doesn't matter," says one federal executive. Adds a management consultant: "It's almost a matter of taste. Do you like chocolate or vanilla?"
What does matter, federal managers and employees say, are clear expectations, regular communication between supervisors and subordinates, and a formal rating system that everyone understands and views as fair and valid. "This is where most systems fall down," says Ken Boxer, president of Strategic Partners, a Potomac, Md.-based training consortium. "People simply don't understand the scale."
Federal evaluation systems often fail to meet managers' and employees' simple criteria. Workers don't trust managers to make fair assessments of their performance. Managers balk because giving low ratings requires vast amounts of paperwork. Also, they don't want to deny their star employees raises. "The way pay is pegged to the scale can have a lot to do with how it gets distorted," says a management consultant. "If employees are on a five-level scale and they need a level-four rating to get a raise, most employees will get a level-four rating."
In an ideal system, no surprises would await employees at evaluation time. With clear goals and measures laid out and with regular discussions with their managers throughout the year, employees would know going into rating meetings whether they were stellar, meeting expectations, or not cutting it. The formal evaluation is just the end product of a manager's yearlong interactions with employees.
It's tempting to think that answering the question of how many rating levels to use will solve evaluation problems, but, in the end, federal managers and employees say there is no scientific answer that will save the day. Managers must have the guts to be honest with their employees. "You have a range in most offices from those who will do anything necessary, think on their feet, be pro-active and support you to those who will do the absolute minimum and bellyache if you ask them to even think pro-actively," says one federal manager. "You must have the wherewithal to rate them as they are."