Tipping the Scales: Behavioral Drivers of Gratuity Decisions Across Restaurants and Rideshare Services

Abstract

This paper examines how default options and group dynamics shape consumer tipping behavior across restaurant and rideshare contexts. Using a novel dataset of over 24,000 transactions from Princeton-area restaurants and 13.8 million NYC taxi rides, we estimate OLS, kernel density and multinomial logit regressions to isolate the effects of default structure and group size on tipping outcomes. We find that gratuity behavior clusters tightly around suggested defaults and that higher default values lead to higher average tips, conditional on tipping. However, in low-service environments, we also show that higher maximum defaults increase the likelihood of tipping omission, suggesting a nonlinear cost to aggressive prompting. Separately, we find no evidence of social loafing, and in fact find a "reverse loafing" effect. Unconditional tip percentages increase significantly with group size across both datasets, contradicting canonical theories of diffusion of responsibility. Overall, our findings extend prior literature by documenting behavioral clustering around tipping defaults, identifying conditions under which interface design increases tip omission, and showing how the core behavioral frameworks of anchoring and social loafing interact with context to influence tipping decisions. We conclude by cautiously offering calibrated policy recommendations for local Princeton restaurants seeking to optimize tip menus, and provide empirical guidance on how behavioral design can enhance employee earnings without triggering consumer resistance.