Is cash king for sales compensation plans? Evidence from a large-scale field intervention

The pervasive use of merchandise (i.e., non-cash) incentives in sales compensation plans is an empirical and theoretical puzzle given the supposed superiority of cash incentives in the standard theory (i.e., principal-agent models) and the scant, and contradictory empirical evidence. We conducted a large scale field intervention that switched 580 salespeople at a large frozen food manufacturer away from their cash plus “ merchandise points” bonus to a commensurate all-cash bonus. After controlling for salesperson, seasonality, year, and target effects, we estimated that sales, on average, dropped by 4.36%. Further, we estimated individual-level sales changes and effort changes to validate our incentive-effort-sales causal chain. Our results show that the top salespeople experienced the largest drops. A post-intervention survey of social and individual difference variables reveals that salespeople from households with more discretionary financial resources, and those who think more abstractly about the uses of cash income exhibited smaller reductions in effort and sales. While the absence of a control group prevents us from making strong causal inferences, this set of results nevertheless provides descriptive and suggestive evidence for separate mental accounts as the most promising explanation for the greater utility provided by merchandise incentives.