A Rand Corp. report produced to guide future U.S. Air Force program plans has concluded that the F-35 Joint Strike Fighter program will cost more than three single-service programs would have done. That conclusion drew a sharp riposte from Lockheed Martin, which accused the report's authors of using “outdated data” that overstated the F-35's projected operating costs by a factor of two.

Lockheed Martin based its criticism on numbers that cannot be found in the report. The company declined to give a source for those numbers, stating that they were “government data.” The Joint Strike Fighter program office distanced itself from the argument, saying it had “no real issues” with the report, and did not confirm any of the company's figures.

Rand's Project Air Force team produced the report, which was requested in 2012 by then-commander of Air Force Materiel Command Gen. Donald Hoffman, as it became clear the JSF would be running many years behind the schedule that was planned up to 2010.

The study was based on historic data up to November 2011, including the fiscal year 2010 selected acquisition report (SAR). Rand, a think tank founded by the Air Force and still closely associated with the service, did not use the fiscal year 2011 SAR (issued in March 2012) which disclosed a three-year slip in development and actually reported higher cost projections than the 2010 report.

Because the JSF program is incomplete, and because no other joint fighter program has been completed as planned, the researchers used data from a variety of programs—from the F/A-18E/F and F-22 fighters to the T-6A turboprop trainer and E-8C surveillance platform—to gauge the historical cost increases in joint and single-service programs.

They did not focus on absolute costs, but on the percentage growth of estimated costs between the launch of a full-scale development program (Milestone B; MS B) and points five and nine years after MS B, the latter corresponding to the most recent JSF data available in late 2011.

The first conclusion drawn from this data was that the average estimated cost of all joint aircraft programs grew faster than that of single-service programs at both the five- and nine-year points, and that the lowest cost-growth rate of any joint program was higher than the fastest-growing single-service program: Even the T-6 grew faster than the C-17, which came close to being canceled because of its cost growth.

The researchers used historical data to estimate production cost saved as a result of joint programs: that is, the benefit of lower unit costs over a longer production run. They found that the maximum benefit for an ideal two-service program—where two services acquire equal numbers of 100% common aircraft—was a 13% cut in unit production cost and 20% in overall acquisition (assuming that production costs were four times research and development costs).

But these theoretical savings were more than offset by the greater cost increase rates observed in practice, making the joint program more costly for both parties. Also, even this calculation did not take into account the higher production and development costs that would be incurred if the partner services procured different versions.

The second part of the Rand report applied this learning to the F-35 program as compared with three notional single-service programs. The MS B baseline R&D cost of three separate programs was based on the JSF program's projections that its R&D cost would be 60% of three single-service programs.

Rand estimated the baseline procurement cost of the separate programs on assumptions that favored JSF. No credit was taken for commonality between the three separate aircraft. Likewise, the MS B baseline operations and support cost assumed zero commonality between the three aircraft and eliminated all economies of scale.

Researchers compared the actual growth of F-35 estimates at the nine-year mark with growth rates for three separate programs based on historic growth with the F-22, the most comparable single-service fighter program. The same adjustments were applied to O&S costs, although a later and higher estimate of F-22 operational costs (at 14 years after MS B) was also included.

The study's conclusion: The JSF estimated life-cycle cost (LCC) in 2010 was already higher than that of three single-service programs. “Under none of the plausible conditions that we analyzed did JSF have a lower LCC than the notional single-service programs.” The report does not recommend any changes to the JSF program, but advises the Air Force to avoid joint projects in future.

Lockheed Martin claims in response that the Rand numbers are outdated and that “the U.S. government stated the O&S [operations and sustainment]costs for the program are approximately $782 billion for a 55-year-long program—not $1.5 trillion as cited in the Rand report.”

The $1.5 trillion figure occurs nowhere in the report, which cites lifetime F-35 O&S as $487 billion in 2002 dollars. This is based on the number in the fiscal year 2010 SAR, adjusted to compensate for 409 Navy/Marine aircraft eliminated from the program.

Neither is Lockheed Martin able or willing to provide a source for its $782 billion figure. The 2012 SAR estimated those costs at $617 billion in 2012 dollars, or $1,113 billion in then-year dollars for operations beyond 2065. That is the closest to $1.5 trillion of any published figures.

More recently, program office director Lt. Gen. Christopher Bogdan was quoted as saying the then-year estimate had been revised downward to $857 billion. One industry source says the revision is largely or entirely the result of changed inflationary assumptions, in which case it would have no bearing on the Rand report because it uses base-year dollars. A JSF program office official agrees that “a lot of the change is inflationary” and says the office is still working with industry to reduce O&S costs.

Lockheed Martin's assertion that the Rand data are “outdated”—echoed in a Pratt & Whitney statement—may not be in the contractors' interest. The Rand report uses fiscal year 2010 SAR data that are actually 10% lower than the most recent official numbers.