There are two ways to be generous with employees. One hands over money and asks to be trusted about where it came from. The other hands over money and the ledger it came from, and invites people to check the arithmetic. The first is an allowance. The second is a stake. Almost everything that separates paternalism from respect at work lives in that difference.
The Semco arithmetic
The working demonstration comes from Semco, as Ricardo Semler has told it. Each unit paid a fixed slice of its own profit to the people who had produced it, and the books were open so anyone could verify the calculation. Two details make the design unusual. First, the share was a rule, not a gift: nobody upstairs decided each year whether the workers had been good. Second, verification was part of the deal, not an insult to it. The observable side effect was the kind managers spend fortunes chasing: people turning off lights, because the money being wasted had become their own. Semler’s compression of the mechanism: nobody guards a pizza. What you cannot verify or influence, you do not protect.
He even proposed an interview question that turns the design into a due-diligence tool: ask a prospective employer what percentage of last year’s profit went to non-executives. A company with a rule can answer in one number. A company with a bonus “philosophy” will need a paragraph, and the paragraph is the answer.
What the evidence supports
The research on shared rewards backs the structure, with an important condition attached. A meta-regression aggregating hundreds of estimates (Doucouliagos, Laroche, Kruse and Stanley, 2020) finds profit sharing positively associated with firm productivity on average, and stronger where it is combined with employee participation in decisions. An earlier review by Kruse and Blasi of dozens of studies reached the complementary conclusion from the ownership side: giving employees a financial stake, by itself, does not even raise their perceived participation; attitudes improve when the stake is combined with actual voice. The money and the say are complements. Pay without voice is an annuity; voice without pay is a suggestion box. Semco’s design bundled them, which is presumably why it produced light-switch behavior rather than a thank-you card.
Transparency itself carries the same design-dependence. Cullen’s review in the Journal of Economic Perspectives (2024) concludes that pay transparency is neither good nor bad in the abstract: its effects depend on what is disclosed and how, with different consequences for comparisons among peers, toward bosses, and across firms. That nuance matters here, because open books at Semco were not a leak of everyone’s paycheck; they were the disclosure of the unit’s performance and the formula tying it to the payout. The transparency was aimed at the deal, not at the neighbors.
The paternalist alternative
Now describe the standard arrangement in the same cold terms. Results are confidential, targets are set above, and once a year a discretionary bonus arrives with a note about a challenging environment. Every element communicates the same thing: you are not a party to this business, you are a dependent of it. Paternalism is often kindly meant, and it economizes on difficult conversations. But it makes a specific bet, that people contribute more when shielded from the truth, and both the Semco record and the meta-analytic evidence bet the other way.
The respect in open books is not the warm glow of inclusion. It is the burden. Reading a P&L, discovering the margin is thinner than the town assumed, watching your own line item appear as a cost: none of that is pleasant, and all of it is what being an adult party to an enterprise feels like. Companies that spare employees this burden usually spare themselves the sharing rule too, which suggests who the protection was really for.
There is also a durability argument. Discretionary generosity is repealed in the first hard quarter, usually with the phrase “difficult decisions.” A formula with open books is much harder to repeal quietly, because everyone can see both the number and the moment it changes. Rules survive moods; that is what they are for.
Generosity, it turns out, is easy to fake and arithmetic is not. A firm that wants to know which one it practices can apply the Semler test to itself tonight: name the percentage, show the books, and see whether anyone starts turning off the lights.
Sources
- Doucouliagos, Laroche, Kruse & Stanley / British Journal of Industrial Relations (2020). Is Profit Sharing Productive? A Meta-Regression Analysis. Profit sharing is on average positively associated with productivity, strongest when combined with employee participation in decisions. https://onlinelibrary.wiley.com/doi/abs/10.1111/bjir.12483
- Cullen / Journal of Economic Perspectives 38(1) (2024). Is Pay Transparency Good? Transparency's effects depend on design: horizontal, vertical and cross-firm disclosure produce different, sometimes opposite, outcomes. https://doi.org/10.1257/jep.38.1.153
- Kruse & Blasi / NBER WP 5277 (1995). Employee ownership, employee attitudes, and firm performance. Reviews with meta-analyses: ownership alone does not raise perceived participation; attitudes improve when a stake is combined with voice. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=225341