![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
This paper describes recent research on cost indexes by the Institute for Defense Analyses. It was performed at the request of the Cost Assessment and Program Evaluation directorate in the Office of the Secretary of Defense to assist in meeting the requirement in the 2009 Weapon Systems Acquisition Reform Act to assess and update the cost deflators the Defense Department uses to adjust for price growth in costing and budgeting major systems. The paper’s focus is on aircraft procurement. The research analyzes deflator algorithms and data to determine the source of the wide differences in aircraft cost growth rates calculated by (a) the Gross Domestic Product deflator for the entire US market basket of goods and services, (b) the national defense index for military aircraft published by the Bureau of Economic Analysis, and (c) the Producer Price Index for civilian aircraft published by the Bureau of Labor Statistics. The study demonstrates an alternative hedonic approach for calculating price indexes by using regression analysis to relate aircraft investment cost to the aircraft’s specific physical and operational design features such as weight and speed.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Flyaway cost, as defined in P-5 Cost Analysis exhibits, is the sum of recurring and non-recurring procurement cost net of support. Recurring cost, by far the major component of total flyaway, is the sum of airframe (contractor furnished), electronics (contractor and government furnished), engines and engine accessories, armament, other government furnished equipment (GFE), and engineering change orders.
2 NAVAIR uses Global Insight estimates for labor and material cost increases to estimate indexes for airframe, engine, and electronics that are combined into an index of flyaway cost for fixed-wing naval aircraft.
3 ‘Defense Budget Materials,’ http://comptroller.defense.gov/budgetmaterials/budget2014.aspx.
4 Constant dollar affordability caps are now also used in mandatory Affordability Analyses at Milestones B and C.
5 A chain-weighted index is considered to be a more accurate inflation gage than the traditional fixed-weighted index because rather than merely measuring periodic changes in the price of a fixed basket of goods, it accounts for the fact that consumers’ purchasing decisions change along with changes in prices. ‘Chain-Weighted CPI,’ Investopedia, http://www.investopedia.com/terms/c/chain-linked-cpi.asp.
6 ‘National Data: National Income and Product Accounts Tables,’ http://www.bea.gov/iTable/iTable.cfm?ReqID=9&step=1#reqid=9&step=3&isuri=1&903=4a.
7 See Footnote 3.
8 Bureau of Labor Statistics, http://www.bls.gov. PPIs are published for both industries such as aircraft production and commodities such as metals and metal products.
9 The annualized growth rate of an index I between 1985 and 2012 is calculated by:
.
10 Although we will be discussing cases where system prices are rising, the deflator algorithms can equally well handle situations where system prices are falling due to deflation or rapidly emerging new technology such as computers.
11 The adjustments are handled explicitly in the model discussed in Section 4.
12 The existing aircraft’s initial price is in the year prior to introduction of the new aircraft. The conversion to contemporaneous dollars is needed in cases where the learning-adjusted price of the new aircraft is estimated by the future-year price of the 100th aircraft when it is delivered. The ‘100 unit method’ (our terminology) was formerly used by BEA (as mentioned in Ziemer and Kelly, Price Measurements) and is also the procedure used in the simulation study of the BEA algorithm carried out in Section 4.
13 By way of terminology, BEA uses the term ‘quality-adjusted base year price’ for the denominator of the deflator – the base year price P1 times the Fi product. The deflator is thus the contemporaneous price Pk divided by the quality-adjusted base year price.
14 The costs for the aircraft deflator are measured at the time of delivery. The deflator for ships is based on yearly progress payments because of their longer production times. Yearly progress payments for aircraft are recorded as additions to inventory and zeroed out at the time of delivery.
15 Recurring flyaway costs are the sum of air vehicle costs (integration, airframe, propulsion, avionics, armament, and engineering change orders) from the Lockheed Martin CDSR for the F-22A (Lot 8, FY 2011) and engine costs from the Pratt and Whitney (United Technologies Corporation) CDSR for the F-119 engine (Lot 6, FY 2005).
Total flyaway adds in nonrecurring flyaway costs, total production includes support recurring and nonrecurring costs, and total overall includes fixed costs for G&A, Miscellaneous, Undistributed Budget, MR, FCCM, and Profit (Fee).
16 Exhibits P-1, P-5, and P-40 in Procurement Programs and DoD Budget Justification Books, all listed in ‘Defense Budget Materials,’ Under Secretary of Defense (Comptroller).
17 See Footnote 3.
18 Military aircraft are described by Mission-Design-Series. For the F-14A, for example, the mission is fighter (F), the design is 14, and the series is A. The aircraft in the first column of Table are new designs, with the exception of the F/A-18E, which was a major change from the previous F/A-18s, and the three F-35 variants, which are being built for different missions and produced in parallel. The aircraft in the second column are either new series (e.g. F-14B) or new block upgrades (e.g. the F-14A + and the F-16C 25/30/50).
19 Higher production rate leads to higher lot quantities, so that the fixed costs per lot are spread over more units with a resulting decrease in the unit fixed costs per lot. Fixed costs for each aircraft program were estimated as a function of peak estimated variable costs which are determined by the quality variables and peak production rate.
Learning is incorporated in the regression by assuming that production of the kth aircraft program in lot t, for example, is increased by the production of the other programs in lot t: , Qtk is the cumulative number of aircraft of program k produced through lot t,
is the cumulative number of aircraft of program k produced up to but not including lot t, qtk is the number of aircraft of program k produced in lot t. There are two special cases where multiple aircraft models have common elements and are produced in the same factory. The resulting shared learning is captured in a third additive argument (
portraying learning spillovers between the relevant models. The regression called for two different spillover parameters λ: one for the EA-18G and F/A-18E, and one for the F-35A, B, and C models.
20 The dummy variables indicate when a model change occurs within a program that results in a loss of learning.
21 In the learning curve, , so if log2 S = −0.25, S = 0.84.
22 The annualized growth rate in percentage terms between index values of I1 and I41 in years 1 and 41 is.
.
23 As Benkard and Bajari explain selection bias in terms of product exits, we chose this as our discriminator; we also found similar results when we analyzed years with market entries.