Promoting Best Practices in a Multitask Workplace: Experimental Evidence on Checklists

Employers often identify best practices and encourage workers to use them. We consider how best to incentivize best practices by analyzing data from field experiments at an auto repair firm. The best practice we consider is the use of checklists by mechanics during car inspections. We find that low-powered incentives to use checklists generate much better outcomes than high-powered incentives. High-powered incentives for checklist use had the effect of crowding out time for actual repairs. We explain this result using a modified multitask principal-agent model that predicts an inverted-U relationship between incentive strength on one task and overall output. These results illustrate how moderate incentives for best practices may have large benefits even in a multitask setting, while strong incentives can be counterproductive.