Shoppers rewarded a fictional retailer that disclosed its labour shortcomings

Two effects run in opposite directions here. Disclosing a shortcoming lowers how fair consumers judge a company to be. Describing the process realistically raises trust. The net depends on which of the two is bigger. Prisca Brosi of Kühne Logistics University, a co-author, put it this way: "While the disclosure of shortcomings leads to lower perceived fairness, this is offset by a positive effect." ^[7]

phys.org did not report effect sizes, per-condition sample sizes, or whether the mixed-disclosure study also included a no-information arm ^[13]. Split evenly between the two studies, the roughly 780 participants work out to about 390 per study, and fewer than that in any single condition ^[12].

Using an invented brand is the strongest part of the design ^[4]. Everyone in the sample came to the company without a prior opinion, so the disclosure wording is close to the only thing varying, and the estimate is not contaminated by whatever a real retailer has done or been accused of. It is also the ceiling on what the result can be read to cover: a real fair-wage report gets read next to the firm's own record.

The all-positive condition was the uninteresting half. "It's not surprising that publishing positive information leads to positive effects, as it signals fairness and trust," Brosi said ^[14]. The comparison the advice depends on is mixed disclosure against silence, and the researchers' recommendation is that publishing mixed results beats publishing nothing ^[15].

The scored outcomes are perceived fairness, trust and word-of-mouth ^[11]. Those are ratings and stated intentions from someone reading a product page. Brosi said companies fear that negative information leads to public criticism, declining sales or regulatory scrutiny ^[9]. Of those three, public criticism and regulatory scrutiny come from audiences other than shoppers.

The statute in the background requires companies to investigate wage and treatment questions and publish the results, though not down to the last detail ^[10]. That discretion is where the internal argument happens. The study covers only one of the audiences reading the output.

For a consumer-facing disclosure page, the evidence supports publishing a mixed record over publishing nothing, and I would act on it there. For a statutory report, the sample left out the audiences that produce the costs Brosi lists. Brosi's recommendation comes with no such condition attached: "Companies should have the courage to publish mixed results," she said ^[8].