Edward Warren was shocked when he learned that the airmen in charge of the nation's nuclear-tipped missiles regularly cheated on tests.
In 2009, Warren was fresh out of the Air Force's Reserve Officers' Training Corps. He had just finished training to become a missile launch officer when he was pulled aside.
"One of my instructors said, 'Hey, just so you know, there is cheating that goes on at the missile bases,' " Warren recalls. "I was repulsed. I thought, 'This can't be, this is terrible.' "
But while serving at F.E. Warren Air Force Base in Wyoming from 2009-2013, Warren saw lots of cheating. The cause, according to Warren and other former missile launch officers reached by NPR, was a culture driven by constant demand for perfection.
Promotions hinged on perfect test scores, and young officers had a choice, he says: "Take your lumps and not have much of a career, or join in with your fellow launch officers and help each other out, and that is what most people did."
This month, the Air Force is scheduled to release the results of its investigation of cheating at another missile base: Malmstrom Air Force Base in Montana.
Already, Air Force officials have publicly stated that 34 officers have admitted to cheating, and dozens more knew about it. The Air Force has since investigated testing at the two other bases where nuclear missiles are kept F.E. Warren and Minot Air Force Base in North Dakota but they have not yet released the results.
In January, Air Force Secretary Deborah Lee James told reporters she thought the cheating ran deep: "We do have systemic problems within the force," she said.
Interviews suggest cheating may have been widespread for years in the missile forces.
NPR reached eight former missile officers, including Warren, who served over decades. All but one admitted that they had participated in some kind of cheating on tests. What's more, they described a culture of cheating that permeated the remote bases that stand guard over the nation's nuclear stockpile.
Being a missile launch officer is a grueling job. Warren and his deputy would regularly drive out onto the windswept planes of Wyoming, Colorado and Nebraska. Often they'd pull off the highway onto a remote road to what might be mistaken for a little ranch house. But beneath it was a fortified nuclear bunker crammed with communications equipment, a toilet and a bed.
Warren and his second-in-command spent 24 hours below, in direct control of 10 nuclear tipped intercontinental ballistic missiles. Their job basically came down to this: wait for a launch order from the president of the United States. And if it ever came, launch the missiles. Fast.
"Very fast. The actual number is classified," he says. (Other former missileers told me they could launch within about a minute.)
It's a job where mistakes aren't tolerated, and everything must be done by the book. There really is a book, several actually, filled with hundreds of checklists. Checklists for everything, from launching a nuke, to letting a maintenance crew into a missile silo, right down to getting lunch.
"Whatever you did, whatever action you were taking, you had to be open to the correct checklist," he says.
Missileers are tested three times a month to make sure they know their checklists. To sit in the bunker, to be in charge of the weapons, you have to get better than 90 percent on every test.
But Warren soon discovered that the tests are used in another way. Because this is a job that everyone is supposed to do in exactly the same way, the tests became a way for the leadership to decide who got ahead.
"It was pretty obvious that if you wanted to succeed, you wanted to move up, you had to meet that near perfection, you know 100 percent average, as close as you possibly could to that, or you wouldn't get promoted," he says.
Cheating was especially common for those just joining the missile forces. The so-called Emergency War Order tests, designed to check whether missileers knew when to launch their weapons, were fiendishly complex. Messing up a test could derail a career.
"Most of the time what it really involved was just the senior launch officers looking out for the more junior launch officers, maybe checking their answers before the test was handed in and saying, 'Hey, look out for No. 5 or 11,' " he says.
Warren counts himself among the cheaters. "I looked out for the more junior launch officers when I was a commander; I made sure they didn't fail, and I received similar help when I was a young officer," he says.
Others went further by doing things like hiding answers in their flight suits, or looking over the other guy's shoulder. At Malmstrom, the 34 officers stand accused of sending and receiving answer sets as text messages.
Most officers NPR spoke to agree that cheating was seen as necessary in a culture that demanded perfection. "Everybody I know that cheated did so to survive," says Brian Weeden, another former missile launch officer. "Given a choice, any other choice, I don't know of anybody who would have done it."
Weeden, Warren and others also agreed that most really do know how to do their jobs. Every month, missile officers are also thrown into simulators and observed by instructors. There's no cheating, and they still get it right.
After he finished his tour, Edward Warren left the Air Force. He was proud of his service, but he was tired of having to cheat to get ahead. He's hopeful the Air Force will make changes.
The simplest, he says, is to stop using test scores to determine promotions. More generally, he says the missile forces must realize that its officers will make an occasional mistake in reading through the hundreds of checklists they have to follow. If they do, there are still plenty of backup systems that will protect against an accident.
Accepting mistakes in a culture of nuclear weapons may sound unacceptable, but Warren says the current system is worse. "Right now what we're doing is setting a standard of perfection, of 100 percent all the time, and no one can achieve that," he says. "A perfection standard is no standard at all."