Would an “honesty pledge” really make employees act more ethically? A study from IESE Business School suggests it might—if worded carefully.

Involving 21,500 participants from the UK and US, the study tested 21 different types of honesty pledges in a tax-evasion simulation. Here’s what they found:

Specific Language Works Best: Pledges that explained exact procedures or clarified gray areas had the most impact on reducing dishonesty.

Typing vs. Ticking: Participants who typed out the pledge showed slightly better honesty than those who just checked a box, though the difference was small.

Moral Appeals Don’t Help: Pledges that appealed to moral values or said honesty helps society had little effect on compliance.

Peer Pressure Works: Pledges that highlighted peers’ honest behavior were more effective in reducing dishonest reporting.

The Overall Impact: Adding a pledge increased compliance by an average of 3.9 percentage points, with the top-performing pledges boosting it by up to 8.5 points.

Of the 21 pledges tested, 10 improved compliance by 4.5 to 8.5 percentage points, with the most effective nearly halving tax evasion.

Demographic Patterns: US participants, especially men, were less likely to comply than UK participants or women.

“Honesty oaths that specified procedures or made situations and behaviors less ambiguous were among the most effective in comparison to control,” the researchers conclude. “Ambiguity and the potential to attribute a dishonest act to the environment are important predictors for individuals engaging in dishonest behavior.”

The findings suggest that well-worded honesty pledges, particularly those that lay out specific rules, can help encourage ethical behavior in the workplace.