OPM’s performance rating system misses the real problem

A long-time federal employee says the new performance ratings system is ‘institutional gaslighting’ and bad for employees and agency mission success.

At the end of this month, my agency will issue its first performance ratings under the Office of Personnel Management’s new appraisal rule. I can tell you now what those ratings will measure. They will not measure performance. They will measure where each of us landed on a curve that has nothing to do with the performance standards written in our appraisals. The curve was drawn by a formula set at OPM headquarters and where you fall on it is close to random.

OPM Director Scott Kupor is right about the underlying problem. For years, roughly 99% of permanent employees were rated fully successful or better, and most of us can name a colleague who coasted while the rest of the team absorbed the work. Accountability for people who are not doing the job is overdue, and career employees know it better than anyone. So let me be clear: Federal performance management needs reform. What is happening now is not reform. Under the new rule, no more than 40% of an agency’s employees may receive one of the two top ratings. The money follows the cap: OPM’s Sept. 21 memo directs agencies to reserve at least 60% of every awards pool for the top tier and to shrink awards for everyone else. Recognition that people earned is now routed to a share of the workforce fixed in advance. A companion reduction in force (RIF) rule makes your last three ratings the main factor in who keeps a job when positions are cut.

The appraisal rule eliminates a mid-level rating, removes the right to grieve the rating you receive and adds “holding employees accountable” as a critical element in every supervisor’s own evaluation. Put those pieces together: Someone must rank at the bottom whether or not anyone is failing. The person ranking you has been told that ranking harshly counts in her own review. You cannot contest it, and where you land decides whether you survive the next cut. Any statistician will tell you that if you force a distribution, you’ll get one. What you will not get is a true picture of performance. What you will get is something worse. When a supervisor is required to put someone in the bottom tier, the biases they carry do not disappear — they decide who goes there.

Decades of research on forced ranking show the bottom tier fills disproportionately with women, older workers, people of color and people with disabilities. Companies that pioneered these systems, including Ford, Microsoft and Goodyear, faced major lawsuits over those exact discriminatory patterns before abandoning them. By removing grievance rights alongside this forced distribution, OPM has eliminated the primary legal remedy for affected employees while manufacturing thousands of defensible discrimination claims. These complaints will flood an already overburdened equal employment opportunity (EEO) process least equipped to handle them.

In the Washington Post, Kupor wrote that merit should mean distinction for those who “knock the cover off the ball.” Baseball keeps more statistics than any sport on earth, and still no club has ever won a pennant by cutting everyone below the team batting average. Pennants are won by the players who move the runner over, cover one another’s range and do the quiet things that never reach a box score. A forced curve tells those players to stop. This model of performance is well known in Silicon Valley, where I spent a decade before federal service and where Mr. Kupor spent his career. It is the 10x model: A few exceptional individuals produce most of the value, management’s job is to find and reward them, and everyone else is overhead. Stack ranking and forced curves are its tools. Some of the companies that pioneered the approach have since abandoned it, having found that the ranking destroyed the collaboration the work depended on.

The federal government is not a venture portfolio. It is closer to the military, and that is no accident. Roughly one in three federal employees is a veteran. The military does not run on 10x individuals. It runs on units, on handoffs and on the quiet competence of people who hold a position for years. That is also how a claims office works, how a Department of Veterans Affairs medical center works and how a disaster response works. In his “Secrets of OPM” blog, Kupor calls that kind of tenure “a lazy substitute for merit.” It is the opposite. It is what keeps the work moving when appointees like him turn over.

Here is what two months under the rule has produced in my agency before a single rating has been issued. People have stopped volunteering. Nobody picks up the ambiguous project, mentors the new hire or sits on the working group, because none of it shows up in a rating. Colleagues who used to share what they know now hold onto it. Morale has plummeted. Supervisors are being asked to pick the people who will fall below the line, and they know what the RIF rule does with that decision. The best people, the ones with options, are updating their résumés, and several have already left. Kupor says the old system drove exemplary employees out in frustration. The new one is doing it faster.

I have been a high performer for decades, with the ratings and awards to show it. Under a forced curve, that means nothing. Someone must fill the bottom, and the math does not care who. For the first time, I find myself wondering whether I am as good as I thought, because the system now needs a certain number of us to be told we are not. That is not accountability. It is institutional gaslighting applied to 2 million people at once.

The people who wrote this rule do not live under it. Political appointees are exempt. Kupor’s own public financial disclosure estimates his assets between $182 million and $573 million. A GS-9 single mom with two kids and a mortgage can lose her job on a rating she cannot contest, from a curve she did not earn. The man who drew the curve will never be rated under it, and if he were, it would not change a thing about his life.

Years ago, frustrated with my team, I went to my supervisor, a decorated combat veteran. He responded by telling me a story about Donald Rumsfeld. In December 2004, at a base in Kuwait, a National Guard soldier stood up at a town hall and asked the Secretary of Defense why his unit was digging through scrap heaps for metal to armor their trucks before driving into Iraq. Rumsfeld’s answer followed him for the rest of his career: “You go to war with the Army you have, not the Army you might want.” The people in the building are the people who will carry out the mission. You cannot rank your way to a different workforce. You can only make the one you have afraid, divided and smaller.

There was a version of this reform that career employees would have embraced: real backing for supervisors dealing with the colleague who is not performing, a faster and fairer way to move that colleague out and the rest of us left alone to do our jobs. Instead, OPM put the rule out for comment, heard from over 600 people who mostly opposed, yet they pushed it through anyway. A better version still exists, and it would cost nothing but the two things OPM’s leadership has yet to show: the curiosity to ask the workforce what it needs, and the humility to listen when it answers.

This column was written by a career federal employee who asked that their name be withheld out of concern for retaliation.

Copyright © 2026 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.

Related Stories

    AP Photo/David J. PhillipFILE - The injectable drug Ozempic in Houston on July 1, 2023. (AP Photo/David J. Phillip, File)

    Expect higher prices and stricter approval requirements for weight-loss GLP-1s in FEHB next year

    Read more
    Getty Images/amgunWarning icon on a digital LCD display with reflection

    AI penetration testing cannot replace independent cybersecurity assurance

    Read more