The History of Performance Management

By Victor Hoang · Updated Oct 24, 2026

In the early 1800s, a Scottish mill owner hung a colored block beside every worker. It showed how well that worker had behaved the day before.

Performance management in the modern sense took shape between 1917 and 1950. Rating scales spread from the US Army into industry and the federal civil service. The practice of recording what workers did and judging officials is thousands of years older. The throughline is a single argument that has never been settled. Is a rating a tool for the person being rated, or a tool for the organization doing the rating?

That argument was already visible in the mill. It runs through the annual review, the forced curve and, since 2012, the move toward lighter check-in conversations. Each step solved a real problem and created the next one. For the practical method today, read our performance review framework. For the wider backdrop, start with the history of management.

Timeline at a Glance

The Tally Keepers (around 1250 BCE-1800s)

Scribes log absences, and Chinese officials are judged on results

Around 1250 BCE, a scribe covered a limestone slab about 38 by 33 centimeters in black and red ink. It is a roster, most likely from the village of Deir el-Medina on the west bank of Thebes, where workers who built the royal tombs lived. A University of Hamburg centre for manuscript studies describes it: a worker's name on the right, then the dates he did not appear, with the reason written in red above the dates. It dates to the 40th year of Ramses II, and the scribe who kept it never signed it. By far the most common reason for staying home was illness.

China shows the next step: consequences. A 2023 study in Frontiers of History in China, based on bamboo and wooden slips from the Han frontier, finds that "incompetence" in a Han official came to include poor performance assessment results. Dismissal was the most common treatment. And the official's superiors or recommenders could be held accountable too.

Then came selection. Under the Sui dynasty (581 to 618), Mark Cartwright writes, the state put greater weight on abilities shown in examinations held in the capital than on family connections. The examinations ran in some form for a thousand years, and the Qing abolished them in 1905. Cartwright reports that only about 1 percent of examinees passed, and that by the end of the Song dynasty some 400,000 candidates sat the pre-tests each year.

Keep the precursors apart from the concept. Egyptian scribes tracked attendance, Han clerks assessed officials and Sui examiners tested candidates, but nobody called any of it performance management, and none of it involved the person being assessed having a say. What carries forward is the shape: a record, a judgment and a consequence.

What this means for teams today: If your review process produces a record of what happened and a consequence for the person, you are running the oldest version of the idea. The question is whether it also helps the person do better next year.

The Silent Monitor (early 1800s-1916)

A mill owner trades the whip for a colored block

At the start of the nineteenth century, discipline at the New Lanark mills was enforced through verbal and physical abuse. A University of Dundee exhibition record says that when Robert Owen became a partner at the largest cotton factory in Scotland, discipline was poor, and that he introduced a more compassionate way to monitor and improve performance. He called it the Silent Monitor.

The Robert Owen Museum's record explains the device. A small four-sided wooden block was hung next to each worker, each side a different color: black for bad behavior, blue for indifferent, yellow for good and white for excellent. The superintendent turned the block each day to match the worker's conduct, and a daily note went into a "book of character" kept for each department. Owen opposed corporal punishment and wanted a substitute. What he built was the first visible, daily, color-coded rating of individual workers in this story.

A century later, governments met the same problem at larger scale. The US Pendleton Act of 1883 required promotion by merit competition to end favoritism, but, according to the Office of Personnel Management's chronology, it set up no centralized appraisal system. Congress addressed that in 1912, when an appropriations act directed the Civil Service Commission to establish a uniform efficiency rating system for all agencies.

The pairing shows the pull between two purposes. Owen's block was meant to replace punishment with a signal. The state's rating system was meant to make fair decisions about people at scale.

What this means for teams today: A rating that people can see and understand daily is gentler than a verdict delivered once a year. If your feedback arrives only at review time, you have kept the verdict and dropped the signal.

The Rating Scale Years (1917-1950s)

Scott's scale leaves the army and enters every office

Industrial psychologist Walter Dill Scott of Northwestern University is the central figure of this era. According to Eli Cook's conference paper for the Business History Conference, Scott invented and institutionalized the employee rating scale: a 5-point scale on which bosses rated workers on traits such as appearance, loyalty, manner, tact, energy, neatness and personality. After it was applied to hundreds of army officers in the war, Cook writes, the scales took off in the 1920s and 1930s despite constant pushback from unions.

Cook's argument is that rating the worker, not the work, was new. Productivity on a factory floor could be counted in widgets. For white-collar and service jobs, the rater's opinion had to stand in for the count. Scott hoped the scale would work as a mirror for the worker as well as a whip for the company.

Government copied the form. The 1923 Classification Act led to a "graphic rating scale" in 1924, in which a supervisor marked a line for each "service rendered", which the OPM chronology calls effective but unpopular. In 1935 the Civil Service Commission replaced it with the Uniform Efficiency Rating System, which grouped factors under quality of performance, productiveness and qualifications. The Performance Rating Act of 1950 went further. It required every agency to build an appraisal system, with three summary ratings: Outstanding, Satisfactory and Unsatisfactory. Its stated purposes were to identify the best and weakest employees and to improve supervisor-employee relations.

The rating-scale era also overlaps with the age of efficiency engineering. Our history of scientific management covers the shop-floor measurement that rating scales tried to extend to office work.

What this means for teams today: A trait list such as "tact" or "energy" invites the rater's personality into the result. If you use a scale, define each point with observable behavior, and have a second person read the evidence.

The Uneasy Look (1954-1970s)

Drucker offers goals, McGregor questions the rating

Peter Drucker's 1954 book The Practice of Management offered another basis for judging managers: agreed objectives instead of personal traits. Our history of management by objectives tells that story, and we do not repeat it here. What matters for this one is the contrast. A trait scale asks "what kind of person is this?" Objectives ask "what did we agree this person would deliver?"

In 1957, Douglas McGregor of MIT published "An Uneasy Look at Performance Appraisal" in the Harvard Business Review. HBR's reprint page records that McGregor observed appraisal within management ranks becoming standard practice over the previous 20 years, and that "the more the method is used, the more uneasy I grow" over its unstated assumptions. His wider argument about the assumptions managers hold about people is told in our history of Theory X and Theory Y.

Researchers began to test the practice. Writing in HBR in 1965, Herbert Meyer, Emanuel Kay and John French noted in their chapter's opening that people wrote pro and con on appraisal, citing McGregor among others, but that "nobody has done any real scientific testing of it". The paper itself is titled "Split Roles in Performance Appraisal".

In 1978 the Civil Service Reform Act took the opposite direction from the critics, at least in scale. The OPM chronology says it required agencies to develop appraisal systems for all federal employees, based on job-related performance standards, with employee participation in setting those standards. Appraisal moved, in OPM's words, to the center of many aspects of personnel management.

What this means for teams today: Be wary of making one conversation do two jobs. A talk about growth and a decision about pay pull in opposite directions, and people hear only the second.

The Ranking Years (1980s-2000s)

A fixed curve spreads, and ratings lose their meaning

Large employers wanted a way to separate their strongest and weakest people. General Electric, during Jack Welch's years as chief executive (1981 to 2001), is credited with first adopting forced ranking. Writing in 2015, consultant Howard Risher described it in Government Executive as a policy that limits top ratings to 20 percent and requires 10 percent of the workforce to be rated unacceptable and terminated, later adopted by several prominent companies and since abandoned by most.

The federal government met the opposite failure. OPM's history says the governmentwide system was standardized in the mid-1980s to five rating levels, with strict links between ratings and cash awards. Rating inflation grew steadily, and the system lost its credibility. Three committees in 1990 and 1991 concluded that appraisal must be flexible and decentralized and should involve employees in its design. In 1993 the National Performance Review recommended decentralized systems that encouraged employee involvement and a focus on improving performance.

One employer broke from the pattern early. Peter Cappelli and Anna Tavis recount in HBR that in 2002 Brian Jensen told Wharton HR executives that drugmaker Colorcon had stopped using annual reviews, giving instant feedback tied to people's own goals instead. The audience was appalled.

These two opposite responses are the core of the era. A fixed curve forces differences that may not exist. Free ratings drift upward until no one believes them. Both end up with the rating as the product rather than the performance. For a measurement alternative that widens what counts, see the balanced scorecard.

What this means for teams today: If everyone is "exceeds expectations", the scale has stopped measuring. If you must differentiate, decide first what evidence would justify each level.

The Check-In Turn (2010s-present)

Big employers drop the annual rating for ongoing talks

In 2012, Adobe abolished its annual review. Adobe's own account, in a 2013 blog post by its HR chief, says it replaced ranking sessions and labels with lighter-weight Check-in conversations built on ongoing feedback, with no formal tool and no set time of year. It reported saving 80,000 hours of managers' time, and said attrition was down year on year. These are the company's own figures. A follow-up post in 2014 repeated the 80,000-hour estimate.

The next large step was Deloitte. In the April 2015 HBR article "Reinventing Performance Management", Marcus Buckingham and Ashley Goodall wrote that the firm was redesigning its system because its process for evaluating people, and then training, promoting and paying them, was "increasingly out of step with our objectives". Months later, according to Government Executive, Accenture announced it would eliminate annual reviews and forced ranking, while keeping feedback. By 2016 Cappelli and Tavis, in HBR, put the shift at more than a third of US companies having abandoned traditional appraisals, with the focus moving from accountability to learning.

This turn is less settled than it looks. Dropping the rating does not drop the decision about pay, promotion or exit, and the decision has to be made somehow. The newest question is what happens when software joins the process. Our guides on AI in performance management and on performance management when AI does the work cover how AI changes feedback, review drafting and the signals managers can trust.

What this means for teams today: Keep what the check-in movement got right: frequent, specific conversation. And write down how pay and promotion decisions will be made, so that the lighter process doesn't hide a heavier one.

Key People

Key Facts

Key Facts: Performance Management History

• An Egyptian roster from around 1250 BCE records absences by worker, date and reason, with illness the most common cause.
• Owen's silent monitor had four colors: black, blue, yellow and white, turned daily by a superintendent.
• The 1950 Performance Rating Act set three summary ratings: Outstanding, Satisfactory and Unsatisfactory.
• Scott's rating scale was applied to hundreds of army officers in the First World War.
• OPM reports that after the mid-1980s five-level system, rating inflation grew steadily.
• Adobe reported saving 80,000 manager hours after dropping annual reviews.

The Pattern Behind the History

We call this the Ledger-to-Conversation pattern. It is our reading of the timeline, not a research result.

1. A record comes first. Egyptian rosters, Han assessments and Owen's books of character all start by writing down what happened.
2. The record becomes a judgment. Scott's scale and the civil service ratings turned notes into scores for the whole organization.
3. The score becomes a product. Forced curves and inflated five-level scales show what happens when the rating matters more than the work.
4. The conversation comes back. Colorcon, Adobe and Deloitte each tried to move attention from the number to the talk.

What It Teaches Teams Today

• Separate the development conversation from the pay decision, and say which one you are having.
• Define each rating level with observable evidence before the cycle starts.
• Let people write the first draft of their own goals, as Drucker's manager's letter did in the history of management by objectives.
• Treat forced distributions with caution. Most employers that adopted them later dropped them.
• Keep a written record of decisions even when you drop the annual form. The oldest rosters survived because someone wrote things down.
• Pair frequent feedback with an honest account of how pay and promotion are decided. Our performance review framework shows one way.