Pentagon battles a brain drain

When Paul Kaminski completed his graduate work in 1971 with degrees from the Massachusetts Institute of Technology and Stanford University, he started building advanced airplanes for the Air Force. By the time he stopped several decades later, he had played a pivotal role in producing a flock of new weapons, including radar-evading stealth aircraft.

If Kaminski were coming out of school today, chances are he would be going to work for the likes of Microsoft or Google.

In the past 10 years or so, as spending on new military projects has reached its highest level since the Reagan years, the Pentagon has increasingly been losing the people most skilled at managing those projects. That brain drain, military experts like Kaminski say, is a big factor in a breakdown in engineering management that has made huge cost overruns and long delays the norm.

Kaminski's generation of engineers, responsible for many of the most successful military projects of the 1970s and '80s, is aging and fewer of the nation's top young engineers, software developers and mathematicians are replacing them. Instead, they are joining high-tech companies and other civilian firms that provide better pay than the military or its contractors, and greater cachet — what one former military industry engineer called "geek credit."

One measure of this shift can be found at the Air Force: Through a combination of budget cuts, the demands of fighting two wars and the difficulty of recruiting and retaining top engineers, officials said, the number of civilian and uniformed engineers on the core Air Force acquisition staff has fallen 35 percent to 40 percent in the past 14 years.

The downsizing has affected the service's ability "to refresh our aging acquisition work force," said Air Force engineering chief Jon Ogg.

With the Pentagon planning to spend $900 billion on development and procurement in the next five years, including $335 billion on major new weapons systems, according to the Government Accountability Office (GAO), alarm is rising among many in Washington.

At a recent hearing of the Senate Armed Services Committee, the chairman, Carl Levin, D-Mich., said cost overruns on military projects, while long a problem, had "reached crisis proportions," and he called for the creation of an internal Pentagon office to oversee costs.

A recent GAO study of 95 military projects worth $1.6 trillion reported projected cost overruns totaling $295 billion, or 40 percent, and an average delay of 21 months. A prime culprit was often deficient engineering management. By comparison, the study found that the Pentagon's 75 major programs in 2000 were 27 percent over budget and 16 months behind.

"We're having awful problems with the execution of defense programs," said Kaminski, the Pentagon's top acquisition executive from 1994 to '97. "It's absolutely critical to start becoming more efficient, more effective."

Kaminski is devoting much of his time as a private citizen to that goal, leading a high-level task force and visiting university campuses and military contractors to proselytize for better engineering management.

Systems engineer is key

As he and other experts explain it, the central problem is a breakdown in the most basic element of any big military project: accurately assessing at the outset if the technological goals are attainable and affordable and managing the engineering to ensure that hardware and software are properly designed, tested and integrated.

The technical term for the discipline is systems engineering. Without it, projects can turn into chaotic, costly failures.

Increasingly, that has become the case. The loss of government expertise has magnified the difficulties associated with another trend: In recent years, the Pentagon has transferred more oversight responsibility to contractors, who often lack sufficient systems-engineering skill and the incentives needed to hold down costs.

Kaminski's task force, organized by the National Research Council, an arm of the National Academy of Sciences and National Academy of Engineering, comprised 18 military experts, working with the Air Force Studies Board, another high-level group.

Their report scolded the Air Force for haphazardly handling, or ignoring, several basic systems-engineering steps: considering alternative concepts before plunging ahead with a program, setting clear goals for a new system and analyzing interactions between technologies.

Kaminski and other experts see no easy fix, noting that high-level reports and recommendations have piled up over the years with little remedial action by the Pentagon.

Gen. Bruce Carlson, commander of the Air Force Materiel Command, said he agreed with the panel's recommendations.

Responding to questions by e-mail, Carlson said that every Air Force program was now required to develop a systems-engineering plan. In addition, the service has established educational programs for its systems engineers and created a new degree in systems engineering at the Air Force Academy.

Still, the military is hard-pressed to compete with the corporate stars of the high-tech era.

At MIT, a 2007 survey showed that 28.7 percent of undergraduates were headed for work in finance, 13.7 in management consulting and just 7.5 percent in aerospace and defense. The top 10 employers included McKinsey, Google, Morgan Stanley, Lehman Brothers, Bain, JP Morgan and Oracle, but not a single military contractor or government office.

The survey showed that the average annual starting salary in finance and high-tech was more than $70,000, compared with $37,000 at the Defense Department. The average in the military industry was $61,000.

MIT does not have comparable survey data for 10 or 15 years ago, but school officials said the trend is unmistakable. "Google calls me every other week looking for systems engineers," said Donna Rhodes, a systems-engineering expert at MIT.

Security, career concerns

The dean of the College of Engineering at Georgia Tech, Don Giddens, noted an additional factor limiting recruitment of highly trained engineers into military jobs: more than half the engineering doctoral candidates at U.S. universities are from abroad and so are ineligible for most jobs requiring security clearances.

Stuart Kerr, a software developer with advanced degrees in mathematics and electrical engineering, left the military sector in 1999 after 10 years to work for a high-tech company. Kerr said the protracted development time for military projects "amounts to a professional death sentence" for scientists and engineers who want to keep up to date with technological advances.

Kaminski, 66, a graduate of the Air Force Academy, spent most of his career running big military projects, including development of the F-117 fighter and B-2 bomber, the first stealth aircraft. These days he serves as a consultant and is a director of General Dynamics and several other military contractors. He also advises the FBI and the National Reconnaissance Office on technology issues.

"Defense acquisition problems should be the subject of acute concern to Americans," Kaminski said in a recent interview at his home office in Northern Virginia.

"This is an area in which our country has enjoyed a fundamental advantage," he said. "It has been vital to our great economic strength, and our strength in national security. If we don't address the problems, those strengths are going to erode. In fact, they are eroding."