#304 1980 · Semco S/A · Industrial manufacturing

A Brazilian factory owner let employees set their own salaries, and paid less on suspiciously honest numbers than a normal review cycle would have

the problem

Pay reviews turn into a negotiation the employee cannot win, run by a manager who can't see the job from the inside

background

A conventional pay review is an asymmetric negotiation: the employee has partial information about market rates and no information about what colleagues earn, the manager holds both, and the manager's incentive is to keep the number low. Companies attack the asymmetry with pay bands, HR benchmarking software and formal review cycles — all of which still leave the manager as the sole gatekeeper deciding the final number, and still leave employees suspecting they're being lowballed.

Ricardo Semler took control of Semco, his father's São Paulo industrial-equipment manufacturer, in 1980 at 21, inheriting a company of 90 employees and $4 million in revenue run on a conventional command hierarchy — and fired 60% of the top managers on his first day as CEO. Rather than replace their judgment with better benchmarking tools, Semler set out to remove the gatekeeper from the salary decision entirely, in an industry and a country where that had no precedent.

what everyone would do

The standard response to worries about pay-review fairness was to add more structure to the manager's decision — pay bands, HR benchmarking software, formal review cycles — while still leaving the manager as the sole gatekeeper making the final call, addressing the tools available to the decision-maker rather than the underlying asymmetry of information and power between employee and manager.

what they saw

Semler saw that the actual problem wasn't the manager's judgment or the data available, it was that pay decisions happened privately, one-on-one, where the employee couldn't see what the manager saw and the manager had every incentive to keep the number low. Giving employees the same information a manager would otherwise use privately, plus visibility from peers who could catch an inflated claim, made self-interest self-limiting rather than something a gatekeeper had to police.

the move

Employees set their own salaries. They fill out a self-evaluation, then weigh their number against four things everyone in the room can also see: published external market rates, internal pay for comparable roles, what peers doing similar work actually earn, and their own stated financial needs — with the company's pay surveys and comparable salary data made openly available rather than held by management. About a quarter of Semco's workforce set pay this way at any time; because the number is defended in front of colleagues who have the same data and would notice an inflated claim, it functions as public accountability rather than an unchecked request.

why it works

Giving employees full access to market data, internal comparable pay and peer earnings, then requiring them to defend their own chosen number in front of colleagues with that same information, makes an inflated claim immediately visible and challengeable by exactly the people best positioned to catch it. Because the claim has to be justified publicly rather than negotiated privately, an employee has a real incentive to set an accurate, defensible number, since asking for too much invites peer scrutiny a private negotiation with a manager never would. Removing the manager as sole gatekeeper also eliminates the structural incentive a manager has to lowball an employee to protect budget or curry favor upward, so the combination of transparency and peer accountability produces pay outcomes disciplined by public scrutiny rather than by a single interested party's judgment.

the payoff

Semco grew from $4 million in revenue and 90 employees in 1982 to $35 million by 1994 and $212 million with roughly 3,000 employees by 2003, sustaining one of the highest growth rates of any company in Brazil through periods when large multinationals entered its markets and the country went through repeated recessions — including 1990, when Semco itself cut management salaries 40% and gave employees approval rights over every item of company spending rather than imposing layoffs unilaterally.

where it breaks

The mechanism requires genuine psychological safety and enough organizational trust that peers will actually scrutinize an inflated claim rather than rubber-stamp colleagues' requests to avoid confrontation — in a culture that avoids conflict or lets social ties override honest scrutiny, the accountability could fail just as easily as a lenient manager would. It also requires genuinely open, accurate market and internal pay data available to check every claim against; without reliable comparable data, self-set pay has nothing solid to be measured against and could drift arbitrarily. And it depends on the population setting its own pay being small enough for peer review to stay practical — Semco's roughly a quarter of its workforce using the process at any time — since extending the same open, peer-scrutinized process across a much larger or more dispersed organization would strain the requirement that colleagues actually know each other's work well enough to catch a bad-faith claim.

what came after

Semler documented the practice in his 1993 book 'Maverick: The Success Story Behind the World's Most Unusual Workplace', which became a management-school staple and made Semco one of the most cited examples of workplace democracy; the self-set-pay model is still taught today, over three decades on, as a case study in transparency substituting for hierarchy in compensation decisions.