A crisis is the one moment when every company reveals which of its two vocabularies was real. The Brazilian economy of 1990 provided the test conditions: a brutal macroeconomic shock, demand collapsing, cash evaporating. The standard corporate script for that moment has two acts, reassurance followed by a surprise layoff, and it was as available in São Paulo then as it is anywhere now. As Ricardo Semler has told it, Semco played a different script entirely.
The deal
Instead of a layoff designed in private, the numbers went up on the wall and the decision went to the workers. The proposition was explicit: a pay cut now, in exchange for two things, control over the company’s spending and a larger slice of the profit when the recovery came. The workforce took the deal. The recovery came. Semler’s one-line commentary on the episode has outlived it: families do not open their books; companies that respect you do.
Look at the deal’s anatomy, because each part is load-bearing. The disclosure came first: no one can accept a sacrifice they cannot verify the need for, so the books had to be open before the question could even be asked. The sacrifice was priced, a defined cut, not an open-ended appeal to tighten belts. The counterparty got consideration in both directions: power now (control over spending) and upside later (the bigger share). And the whole thing was chosen, which converted the workforce from objects of a crisis plan into parties to one. Every element is the opposite of the standard script, which conceals the numbers, extracts unpriced sacrifice, offers no upside, and calls the result a difficult decision.
Why the design works
Two decades later, economists produced large-scale evidence for the mechanism the deal intuited. Aghion, Bloom, Lucking, Sadun and Van Reenen, studying firms across ten OECD countries plus US establishment data through the Great Recession, found that companies that had delegated more decision-making power to local managers before the crisis outperformed centralized ones in the sectors hit hardest. Their explanation is informational: turbulence raises the value of local knowledge, because headquarters’ models break exactly when conditions leave the range they were fitted on. The authors are careful to qualify that decentralization is not universally optimal; its advantage is specifically a bad-times advantage.
That is the 1990 deal in academic dress. A crisis moves the crucial information to the edges, to the people who know which supplier can wait, which machine can idle, which expense was always vanity. A secret war-room layoff discards that information along with the people who hold it. Semco’s trade, wages for control, moved decisions to where the information already was, and paid for the move with equity in the outcome. The profit-share leg matters as much as the control leg: the meta-analytic evidence on profit sharing finds it most productive precisely when combined with employee participation in decisions, which is the exact bundle the deal offered.
The general lesson
The transferable content is not “cut pay in a crisis.” It is a three-move sequence available to any leadership team before the next bad year. Show the constraint: real numbers, early, ugly if they are ugly. Price the sacrifice: a defined ask with a defined duration beats an open-ended one, because it can be agreed to rather than endured. Share the upside: whoever funds the survival owns part of the recovery, in writing. Companies run the first move at best, and usually none, which is why their crises produce survivors rather than parties, and why the trust destroyed in one February takes years of Decembers to rebuild.
Objections usually arrive in two forms. One says workers will never accept cuts; the 1990 case answers that they will, when the need is verifiable and the sacrifice is compensated with power and upside. The other says disclosure in a crisis is dangerous; but the dangerous version is the leak, not the briefing, and the workforce always finds out anyway, on worse terms and with worse timing.
There is also a quieter lesson about timing. The 1990 deal was possible because the open books predated the crisis; disclosure improvised under duress reads as manipulation. The institutions of honesty have to be built in good times to be spendable in bad ones.
The test of family rhetoric is a bad year. The test of an adult contract is the same year. One of them has a documented case of passing.
Sources
- Aghion, Bloom, Lucking, Sadun & Van Reenen / American Economic Journal: Applied Economics 13(1) (2021). Turbulence, Firm Decentralization, and Growth in Bad Times. Firms that delegated decisions to local managers before the Great Recession outperformed centralized firms in the hardest-hit sectors; mechanism is the rising value of local information under turbulence. https://www.aeaweb.org/articles?id=10.1257%2Fapp.20180752
- Doucouliagos, Laroche, Kruse & Stanley / British Journal of Industrial Relations (2020). Is Profit Sharing Productive? A Meta-Regression Analysis. Profit sharing associates positively with productivity, most strongly when combined with employee participation in decisions. https://onlinelibrary.wiley.com/doi/abs/10.1111/bjir.12483