Assessment of Productive Efficiency of Airports

Abstract

The move towards commercialization and privatization has pressured airports to become more productive and competitive. The need to devise an overall (total) productivity measure is increasingly important in airport business. The dissertation made three major research contributions. First, it assessed the productivity of airports operating in multiple airport systems (MASs). Second, it developed a more complete total factor productivity measure by considering joint production of desirable and undesirable outputs. Third, it developed models for explaining variations in productive efficiency. These are accomplished in two case studies. In case study 1, the Data Envelopment Analysis (DEA) is used to assess relative total productivity of 72 airports operating in 25 MASs during 2000 - 2002. The results indicate that highly utilized airports such as O'Hare International, Los Angeles International, Heathrow/London and LaGuardia are classified as efficient. The Censored Tobit regression model suggests that runway utilization market dominance, proportion of international passengers and ownership can be used to explain variations in productive efficiency. In case study 2, the directional output distance function is applied to assess the productivity of 56 U.S. commercial airports during 2000 - 2003. Delays are considered as undesirable outputs. There are several important findings and insightful implications. First, about half of U.S. airports are actually operated efficiently. These airports include busy airports such as Hartsfield-Jackson Atlanta, LaGuardia, and Memphis together with less busy airports with relatively low delays such as Baltimore/Washington International and Oakland International. Second, the overall system has potential to accommodate about 1,550 million passengers, 26 million movements and 34 million tons of cargo. Third, during 2000 - 2003, annual growth of productivity is modest in the range of -1.3% to +1.8%. Fourth, by ignoring delays the assessment provides drastically different results in terms of number of efficient airports, level of inefficiency, ranking, and estimated maximum possible outputs. Fifth, the consideration of undesirable output is as important as the consideration of additional inputs and desirable outputs. The Censored Tobit regression model suggests that runway utilization, proportion of international passengers and average delay per passenger can be used to explain variations in productive efficiency.