AI Summary & Key Takeaways
AI InsightThe findings show that flight frequency alone cannot serve as a cargo capacity target, and that the load definition, structural zeros, sample and weights must be reported explicitly.
📌 3 Key Takeaways
Abstract
Although flight frequency is a common indicator in cargo planning, the estimand depends on data and model choices. This study builds a panel that follows 52 airports over 39 quarters from 120 monthly releases published by the General Directorate of State Airports Authority (DHMİ) and examines the within-airport relationship between commercial aircraft movements and cargo tonnage alone. The primary estimate is the contemporaneous log-log relationship in active quarters in which both variables are positive. A 1% airport-specific change in commercial movements is associated with a 0.650% change in cargo tonnage. The 95% confidence interval (CI) runs from 0.345 to 0.955, and the model explains 5.3% of the within-airport variation net of fixed effects. The Poisson pseudo-maximum-likelihood (PPML) coefficient, which retains zero cargo in quarters with positive movements, is 0.356; the load coefficient is 0.933 and the cargo coefficient that weights high-volume airports is 0.945. At the median of the active sample, a 10% difference in movements corresponds mechanically to 1.1 tonnes, and at the three high-volume airports to approximately 10,100 tonnes. These calculations are not causal forecasts. The findings show that flight frequency alone cannot serve as a cargo capacity target, and that the load definition, structural zeros, sample and weights must be reported explicitly.