# 1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
ASSESSING THE RELATIVE EFFICIENCY OF AIRCRAFT
MAINTENANCE TECHNOLOGIES: AN APPLICATION OF DATA
MILO W. PECK, JR.
Department of Accounting, Fair®eld University, Fair®eld, CT 06430, U.S.A.
CARL A. SCHERAGA*
Department of Management, Fair®eld University, Fair®eld, CT 06430, U.S.A.
RUSSELL P. BOISJOLY
Dean and Department of Finance, School of Business, Fair®eld University, Fair®eld, CT 06430, U.S.A. (Received 17 October 1996; in revised form 04 March 1997)
AbstractÐThis study focuses on discretionary maintenance strategies and their relationship to aircraft reliability, as measured by the percentage of scheduled ¯ights delayed because of mechanical problems. The methodology of Data Envelopment Analysis is employed to identify the various strategies employed by the major airlines over the time period 1990±1994. Additionally, this methodology allows for a normative assessment as to which strategies are relatively ecient. Furthermore, the speci®c strategies utilized by ecient and inecient airlines can be compared at a micro-level and thus quanti®able recommendations for the latter group can be suggested. # 1998 Elsevier Science Ltd. All rights reserved 1. INTRODUCTION
There has been much concern of late that rising costs and shrinking pro®t margins may have undesirable eects on the ability of airlines to maintain acceptable levels of safety performance. Speci®cally, such concern calls into question discretionary managerial strategies on the part of airline executives with regard to the performance of aircraft maintenance. The researcher is challenged to identify and quantify those discretionary strategies most likely to impact ®rm and industry levels of maintenance eectiveness. Furthermore, once the array of strategies is identi®ed, there exists the additional research challenge of determining, in a meaningful manner, those strategies deemed to be most ecient. Prior empirical work strongly suggests that airline safety has not declined since deregulation. Previous studies have utilized accident and passenger fatality rates in examining whether airline safety performance has deteriorated since deregulation. Rose (1989, 1992) in regressing the log of accident rates on a time trend variable found that ``improvements in airline safety do not appear to have slowed appreciably since deregulation''. She does caution that in the more recent time period, 1987±90, there is a tendency for accident rates to lie slightly above the long-term trend line, possibly suggesting that regulatory eects may operate with long lags. Barnett and Higgins' (1989) calculations demonstrate that the fatality risk for a passenger on a domestic airline declined from an average of 1 in 2.5 million ¯ights over the period 1971±1978 to 1 in 7.4 million ¯ights over the period 1979±1986. If the researcher restricts the calculation to established carriers, the fatality risk over the latter period is 1 in 11.8 million ¯ights. Oster and Zorn (1989) ®nd, in comparing the periods 1970±1978 and 1979±1985, that accident rates due to pilot or controllers' errors, equipment failure and other aircraft declined in the deregulated period. This study focuses on a more subtle issue. The question to be investigated is whether there are signi®cant dierences, for major airlines (part 121 carriers), in the eciencies of maintenance *Author for correspondence
Milo W. Peck Jr. et al.
technologies across airlines. Furthermore, one would also like to investigate the eciency trends over time across airlines. The particular phenomena examined is the relationship between maintenance expenditure allocation strategies and the percentage of scheduled ¯ight arrivals delayed because of mechanical problems. This allows for...