The Pentagon must reform its acquisition process to transition new technologies from the research and development phase into production more quickly. But while the Pentagon needs to restructure, the last thing the Defense Department needs is more bureaucracy.

“We have a problem perennially of connecting the research and development enterprise to the production,” Defense Secretary Ashton Carter told an audience at the Naval War College in Newport, Rhode Island, on May 25. “The lasting in the world we need is bureaucracy.  To me, there is way too much of that.”

The basic problem is that the Pentagon is structured to buy big-ticket items like large warships that might take a long time to build. Unfortunately, given the speed at which technology is evolving, that archaic process does not work well anymore.

“There are some things like capital ships that naturally go on longer time scales, but unfortunately a lot of the reflex of our acquisition system is to operate on that kind of time scale, which as you point out is completely inappropriate when it comes to things like drones; when it comes to things like cyber,” Carter said. “It's the central reason why you see me pushing so hard to connect us with the tech community and rebuild those bridges that were so strong when I started out in this business, between the technical community in the country at large, and the defense world.”

American industry has worked hand-in-glove with the defense apparatus since World War II. However, as technology evolved—particularly since the dawn of the information age—there has been a divergence. Unlike during previous decades where advanced technology mostly originated with the military, the civilian sector has taken the lead in the years since the end of the Cold War.

While this is especially true in the technology sector, even for industrial items such jet engines, the civilian sector is far ahead of the military. Indeed, a typical airliner engine is far more technologically advanced than even the latest military turbines like the Pratt & Whitney F135. “It's also no longer true as it was when I started out that all technology—most technology of consequence originated in the United States and originated in the defense or government environment,” Carter said. “That's just not true anymore.”

Thus the Pentagon has to evolve to engage the civilian sector. But unlike the military, the civilian technology sector operates on timescales of months rather than years—which is a completely foreign concept to the Pentagon. Systems like the F-35 have taken the better part of two decades to reach fruition while entire generations of information technology have come and gone in that same span. As such, cutting edge tech companies just can’t relate to the Pentagon or its needs.

“They don't know us and our problems.  They don't know how to work with us as easily,” Carter said. “So we have to have a different model where we're connected.  That's why you see me out in Silicon Valley, up in Boston, out in Austin.  And so we're trying to get the most vibrant parts of our tremendously innovative country harnessed to the United States.”

To operate in this new environment, the Pentagon will have to reform is processes. But the Pentagon—even as it works to ease red-tape—still has to strike a balance between outsourcing the management of development programs to contractors and doing it in-house.

“That is a difficult balance,” Carter said. “[The U.S. Army’s now defunct] Future Combat Systems was an example where we outsourced the systems engineering, configuration control, everything—and that was the rage at the time.  You know, let industry do it.  They will have the expertise to—well that didn't work.  Almost all the time it failed.”

The FCS model ceded too much of the program management authority to Boeing, which was the prime contractor for that massive family of lightweight ground combat vehicles and the data-networks that held them together. In many ways, the FCS became a poster-child of program mismanagement during the early 2000s. But going towards a fully government managed process is probably not the right answer either.

“The pendulum needs to come back,” Carter said. “Now, not all the way back because I do want there to be shared responsibility and shared expertise in industry and also I need to be careful about what we can take on because—to be blunt about it, our acquisition system is uneven.  Some parts of it are up to that job and other parts, we need to work on and get up to that job.”

Indeed,  while the pendulum has swung back more towards the Defense Department managing its programs, the government still needs to improve its expertise. However, there are also areas where the program management pendulum has swung too far toward government. In some cases the bureaucracy—the government’s acquisition executives—have more or less bypassed the individual military services.

“That's an area also where the pendulum swung too far and the service chiefs basically didn't have anything to do with acquisition for 15, 20 years or so, or not nearly enough,” Carter said. “I think that was the pendulum going too far.  I'd like to see it get back the other way.”

But for the services chiefs to have a more active role in the procurement process, those senior leaders will need more training. “I've told the chiefs in order to do that, you have to educate yourselves because most of you and most of the senior leadership by now in all the services have gotten there not because of their expertise in acquisition,” Carter said. “That was the way it was when I started out in the Cold War.”

Carter pointed out that it would be a challenge for the service chiefs to take a more active role in the procurement process. During the Cold War, there were fewer operational demands on the military since there were few shooting wars. The entire focus was on being ready for World War III—the modern world is quite different.

“Being good at getting ready to do things was what made you proficient in your service, and so acquisition executives who rose to the top, that and there's nothing wrong with this,” Carter said. “But we've had a more operational set of decades, and so we have tend to have more operators at the top and so these guys need to get in the game.”

One thing the Pentagon should not do, however, is split research and development from acquisitions. “If you really want to go after acquisition, you don't take up research and engineering and separate it from acquisition,” Carter said. “That was my point earlier. They've got to come together faster because the future is going to go to those who are agile, creative, up to date, competitive.”

Dave Majumdar is the defense editor for the National Interest. You can follow him on Twitter: @davemajumdar.