Long before 'skills-based hiring' acquired conference panels, a Brazilian industrial company was practicing its blunt ancestor: not looking at diplomas at all. Ricardo Semler has said he never looked at a diploma before hiring anyone. The company's factories, as he tells it, were run in part by engineers without degrees, while MBAs from excellent schools proved unable to run a meeting. The distilled version of the policy fits in a sentence he has used since: trust doesn't come with a gown.
The emblem of the approach was a maintenance technician, no diploma, leading two hundred people. Not because the company lowered its standards, but because everyone trusted him, and in a factory, where output is visible and mistakes announce themselves, trust of that kind is not a feeling. It is an accumulated audit, conducted daily, by the only evaluators with complete information: the people who work beside you.
What the diploma was replaced with
The interesting part of the Semco story is not the absence of the credential but the presence of its replacement. A company that ignores diplomas still needs to answer the question diplomas pretend to answer: who is competent to be given responsibility? Semco's answer, in Semler's telling, was to move the evaluation from before the relationship to inside it, and to hand it to the governed. Everyone rated their boss anonymously twice a year; a manager scoring low twice in a row lost the function, Semler included in the rule. Teams could veto a manager outright. Returning to technical work at the same salary was a legitimate exit, so the system could strip authority without destroying people.
Notice the structure. A diploma is a one-time certification, issued by an institution that will never observe the holder's actual management, valid forever. The Semco instrument was the inverse on every axis: continuous, issued by direct observers, and expiring every six months unless renewed by the experience of the led. It was not a softer standard than credentialism. It was a dramatically harder one, which is presumably why the first year was uncomfortable, and why, as Semler tells it, managers who shouted began instead to ask.
Management research would later formalize why this is not merely charming. Lee and Edmondson's canonical 2017 review of self-managing organizations lists Semco among the documented precursors of radically decentralized management, and the selection literature has long pointed in the same direction on evidence: Schmidt and Hunter's synthesis of 85 years of research found samples of actual work among the strongest predictors of job performance, with credentials far weaker. Watching people work, it turns out, is the methodologically respectable option. Semco simply built a company where watching was the institution.
The uncomfortable generalization
It is tempting to file the story under folklore: one company, one era, one storyteller. The honest reading is narrower and sharper. Semco's practice demonstrates that a mid-sized industrial firm could allocate authority over hundreds of people using zero educational credentials, provided it paid the alternative cost: dense, recurring, sometimes awkward evaluation by peers and subordinates. The diploma is not free either; its costs are merely paid earlier, by someone else, and in a currency (tuition, years) that excludes many of the capable.
The generalization that survives is about where evidence lives. Credentials are portable claims for strangers; they exist because most hiring happens between strangers. Semco reduced its dependence on stranger-evidence by making fewer decisions at the stranger stage: real evaluation happened after entry, continuously, with consequences in both directions. Modern equivalents (probation periods taken seriously, paid trials, internal mobility based on team demand) are all partial rediscoveries of the same move.
The question the story leaves behind
The lasting provocation is not 'stop requiring degrees.' It is a question about any organization's evidence diet: how much of what you believe about your people's competence comes from documents produced before you met them, and how much from structured observation since? A company that cannot answer proudly is running on parchment. Semco's factories ran on the other thing, the slow, compounding, unforgeable credential issued by colleagues, twice a year, forever renewable and forever revocable. There is a name for that instrument. It is called a reputation, and no institution has ever managed to sell one.
Sources
- Lee & Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing (Research in Organizational Behavior 37). Canonical definition of self-managing organizations; Semco cited among documented precursors of radical decentralization. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Schmidt & Hunter (1998). The Validity and Utility of Selection Methods in Personnel Psychology (Psychological Bulletin 124, 262-274). Work sample tests among the strongest predictors of job performance (mean validity .54); education far weaker. https://home.ubalt.edu/tmitch/645/session%204/Schmidt%20&%20Oh%20validity%20and%20util%20100%20yrs%20of%20research%20Wk%20PPR%202016.pdf