Cost Forecasting Models for the Air Force Flying Hour Program

Abstract : The fiscally constrained environment in which the Air Force executes its mission places great emphasis on accurate cost estimates for planning and budgeting purposes. Inaccurate estimates result in budget risks and undermine the ability of Air Force leadership to allocate resources efficiently. This thesis evaluates the current method used by the Air Force and introduces new methods to forecast future Flying Hour Program costs. The findings suggest the current forecasting method's assumption of a proportional relationship between cost and flying hours is inappropriate and the relationship is actually inelastic. Prior research has used log-linear least squares regression techniques to forecast Flying Hour Program cost, but has been limited by the occurrence of negative net costs in the underlying data. This research uses time series and panel data regression techniques while controlling for flying hours, lagged costs, and age to create net costs models and an alternative model by separately estimating the two components of net costs which are charges and credits. Finally, this research found neither the proportional, net costs, nor charge minus credit models is a superior forecaster. As such, the models introduced in this research may be used as a cross check for the current method.

[1]  Patrick D. Armstrong.  Developing an Aggregate Marginal Cost Per Flying Hour Model for the U.S. Air Force's F-15 Fighter Aircraft , 2012 .

[2]  Shu-Ping Hu.  THE IMPACT OF USING LOG-ERROR CERS OUTSIDE THE DATA RANGE AND PING FACTOR , 2005 .

[3]  F. Diebold,et al.  Comparing Predictive Accuracy , 1994, Business Cycles.

[4]  Eric M. Hawkins.  Predicting the Cost per Flying Hour for the F-16 Using Programmatic and Operational Variables , 2005 .

[5]  Eric J Unger.  An Examination of the Relationship Between Usage and Operating-and-Support Costs of U.S. Air Force Aircraft , 2009 .

[6]  Transformation of variables to linear combinations , 1983 .

[7]  Scott Houser,et al.  A Physics-Based Alternative to Cost-Per-Flying-Hour Models of Aircraft Consumption Costs , 2000 .

[8]  Paul Newbold,et al.  Testing the equality of prediction mean squared errors , 1997 .

[9]  Richard G. Harris,et al.  Applied Time Series Modelling and Forecasting , 2003 .

[10]  H. White.  A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity , 1980 .

[11]  Edward D. White,et al.  Empirical Evidence Relating Aircraft Age and Operating and Support Cost Growth , 2008 .

[12]  M. Pesaran,et al.  Testing for unit roots in heterogeneous panels , 2003 .

[13]  Matthew E. Laubacher.  Analysis and Forecasting of Air Force Operating and Support Cost for Rotary Aircraft , 2012 .

[14]  C. Granger,et al.  Spurious regressions in econometrics , 1974 .

[15]  J. Osborne.  Notes on the use of data transformations. , 2002 .

[16]  Van Dyk,et al.  Forecasting Flying Hour Costs of the B-1, B-2, and the B-52 Bomber Aircraft , 2012 .

[17]  Craig C. Sherbrooke,et al.  Predicting Wartime Demand for Aircraft Spares. , 1997 .

[18]  Gregory G. Hildebrandt,et al.  An Estimation of USAF Aircraft Operating and Support Cost Relations , 1990 .

[19]  Gregory G Hildebrandt.  Budget Estimating Relationships for Depot-Level Reparables in the Air Force Flying Hour Program , 2007 .