Airlines showcase their on-time performance (OTP), a globally accepted operational performance metric, to demonstrate punctuality, service reliability, and attract air travelers. Airlines can adopt passive strategies, such as "schedule padding," and active strategies, such as making "operational changes/improvements," to improve their OTP. While there is a high degree of variation in airlines' OTP from year to year, it is unclear if and the extent to which airlines' active or passive actions impact their OTP because of factors, like weather, outside an airline's control. We develop a framework in this paper to study the impact of these active and passive actions on the OTP of airlines. Additionally, we study the effect of these strategies on OTP rankings, routinely used to compare airlines. Our methodology builds on the structural estimation model developed in prior literature and replicates the typical schedule planning process observed in the airline industry. We use an 11-year panel data of flights operated by US domestic carriers from 2005 to 2015 to measure OTP changes, schedule padding, and operational changes. Broadly, active actions include (i) changes to the airline's network structure (e.g., flight routes and schedules) and (ii) other operational changes to improve flight performance (operational improvement). We demonstrate through our analysis that operational improvements have the highest association with both the change in OTP and OTP rankings of airlines, followed by schedule padding; the impact of network changes on OTP and OTP rankings is the lowest. Our framework also accounts for the impact of competition on an airline's OTP ranking. We show that while an airline's own actions can improve its OTP ranking, a competitor's action may negatively affect the ranking. In fact, a competitor's passive strategy of schedule padding may have a higher impact than an airline's own active strategy of changes in network structure. Our results also indicate that the potential impact of operations improvements is the highest for full-service airlines and the lowest for leisure airlines. Furthermore, we show that the impact of operational improvements and buffer adjustments decreases with an increase in the variability of the travel time of a route.
