Aircraft Cost Growth And Development Program Length: Some Augustinian Propositions Revisited - Statistical Data Included

Acquisition Review Quarterly,  Summer, 2000  by Henry L. Eskew

• E-mail
• Print
• Link

This paper examines two notions that were popularized by Norman Augustine. The first is that growth in the cost of successive generations of tactical aircraft is more an inherent (time-driven) characteristic of such programs than a reflection of changes in their technical parameters. The second is that the design and build phase of aircraft development programs has remained virtually unchanged for 40 years, implying that no systematic relationship exists between the characteristics of a program and the length of its development cycle. Models resulting from this examination, which suggest certain modifications to Augustine's original propositions, are tested against recent data from the F/A-18E/F program.

More than a decade ago, Norman Augustine (1986, p. 143) provided a humorous characterization of growth in the costs of military aircraft:

In the year 2054, the entire defense budget will purchase just one aircraft. This aircraft will have to be shared by the Air Force and Navy 3-1/2 days each per week except for leap year, when it will be made available to the Marines for the extra day.

Augustine actually produced a scatter plot and some trend lines to support this prophecy. Elsewhere in the same publication he wrote (1986, p. 140):

...the cost of an individual airplane has unwaveringly grown by a factor of four every 10 years. This rate of growth seems to be an inherent characteristic of such systems, with the unit cost being most closely correlated with the passage of time rather than with changes in maneuverability, speed, weight, or other technical parameters.

Even the most casual observer of the defense marketplace would agree that long-term growth in the costs of tactical aircraft has been substantial. When Augustine says that unit cost has grown by a factor of four every 10 years, which equates to an annual growth rate of 15 percent, he makes no adjustment for the normal increase in manufacturing prices over time--inflation. Other factors having a bearing on aircraft unit costs (besides changes in technical characteristics) are first, the total procurement quantity of a given type and model--the so-called learning-curve effect--and second, the number of units produced in a given year--the production-rate effect.

This article seeks first to disentangle these factors--inflation, technical characteristics, learning, and production rate--from the growth in aircraft costs experienced over a 30-year period (1950-1980). The result constitutes an estimate of the real rate of cost growth, meaning the rate that is associated strictly with the passage of time. That result is then tested against data from a current program, the Department of the Navy's F/A-18E/F, in an effort to see if the same rate of cost growth continues, or--as one would hope--if it has abated to some degree.

Another of Augustine's propositions (1986, p. 356), and one of considerable interest in defense acquisition circles, is the following:

The duration of the design and build phase of aircraft development programs has remained virtually unchanged for 40 years. This period is approximately the same for government projects, commercial projects, and, for that matter, projects undertaken in the Soviet Union.

Based on a scatter plot showing no trend between months-to-first-flight and year-of-first-flight for a combined set of military and commercial aircraft, this statement strongly implies that no systematic relationship exists between the characteristics of an aircraft program and the length of its development cycle. Historical data examined later in this article suggest that the length of a tactical aircraft development program has been systematically related to a standardized measure of the aircraft's eventual procurement cost. We consider the cost measure to be a proxy for program complexity or sophistication. As with the trend in cost growth, we tested that relationship against recent data from the F/A-18E/F program.

TACTICAL AIRCRAFT COSTS

A consistent source of procurement cost and technical characteristics data (McNichols, 1983) was available for 17 fighter and attack aircraft programs. The oldest of those was the F-89. The year in which its first operationally configured production unit was delivered--the measure of time employed throughout the paper--was 1950. The most recent aircraft is the F-18A; its year of first delivery was 1980. The other programs were the A-4, A-6, A-7, A-10; F-4, F-14, F-15, F-16, F-100, F-101, F-102, F-104, F-105, F- 106, and F-111.

Measures of unit flyaway cost--the cleanest quantification of procurement cost--were constructed as follows, with flyaway defined to include airframe, engine, electronic, and armament costs, but not spare parts or other support items. First, to eliminate the learning-curve effect, we focused on the cost of the 100th production unit. Because military aircraft are procured in lots rather than by individual units, the cost of unit 100 can only be approximated. Dividing total annual flyaway cost in the year that included the 100th unit by that year's procurement quantity gives an approximate unit-100 cost in undeflated dollars. Then, to remove the effects of inflation, we applied a procurement-cost deflator (Office of the Secretary of Defense, 1991) to convert the cost measures to constant fiscal year 1990 dollars. Figure 1 is a plot of approximate unit-100 flyaway cost in millions of 1990 dollars against year of first delivery.