Any effort to rebuild TSA and get airport security right in the U.S. has to start with two basic principles:

First, the TSA's mission is to prevent a catastrophic attack on the transportation system, not to ensure that every single passenger can avoid harm while traveling. Much of the friction in the system today results from rules that are direct responses to how we were attacked on 9/11. But it's simply no longer the case that killing a few people on board a plane could lead to a hijacking. Never again will a terrorist be able to breach the cockpit simply with a box cutter or a knife. The cockpit doors have been reinforced, and passengers, flight crews and air marshals would intervene.

Second, the TSA's job is to manage risk, not to enforce regulations. Terrorists are adaptive, and we need to be adaptive, too. Regulations are always playing catch-up, because terrorists design their plots around the loopholes.

I tried to follow these principles as the head of the TSA, and I believe that the agency made strides during my tenure. But I readily acknowledge my share of failures as well. I arrived in 2005 with naive notions of wrangling the organization into shape, only to discover the power of the TSA's bureaucratic momentum and political pressures.

There is a way out of this mess—below, I'll set out five specific ideas for reform—but it helps to understand how we got here in the first place.

The airport checkpoint as we know it today sprang into existence in spring 2002, over a month and a half at Baltimore/Washington International airport. New demands on the system after 9/11, like an exhaustive manual check of all carry-on bags, had left checkpoints overwhelmed by long lines and backlogs. A team of management consultants from Accenture delved into the minutiae of checkpoint activity at BWI: How long did it take to pass from one point to another? How did the behavior of travelers affect line speed? How were people interacting with the equipment?

The consultants had a million ideas for improvement, but with no infrastructure, acquiring even the most ordinary items became a quest. For example, before passengers walked through the metal detectors, they needed to place their keys, jewelry and change into a container. But the long, skinny plastic dishes in use at the time tipped over. So a team member went to PetSmart, bought a bunch of different dog bowls and tested each one. The result was the white bowl with a rubber bottom that's still in use at many airports. (Please, no jokes about the TSA treating passengers like dogs.)

One brilliant bit of streamlining from the consultants: It turned out that if the outline of two footprints was drawn on a mat in the area for using metal-detecting wands, most people stepped on the feet with no prompting and spread their legs in the most efficient stance. Every second counts when you're processing thousands of passengers a day.

Members of Congress, who often fly home to their districts for the weekend, had begun demanding wait times of no longer than 10 minutes. But security is always about trade-offs: A two-minute standard would delight passengers but cost billions more in staffing; ignoring wait times would choke the system.

After I was confirmed as TSA administrator in 2005, one of the first things I did in office was to attend screener training at the Minneapolis-St. Paul International Airport.

I sat down at a computer with Gary, a solidly built guy in his 40s with a mustache and a shaved head. Gary pointed at a screen that simulated the carry-on bag monitors at checkpoints. "What do you see?" he asked, a half smile on his face.

I stared at the series of colorful, ghostly images that Gary froze on the screen and tried to pick an easy one. "Well, that's a computer or some electronic, there are wires, maybe a battery." The sharp edges were easy to pick out, and the recognizable pattern of a motherboard jumped out. "But I don't know about that big orange blob on top of it."

"Right," said Gary. "The orange-colored part…. That means it's organic. Anything made of organic material—clothes, shoes, food—it's all going to register orange here."

As a confidence boost, Gary gave me a series of images with guns and knives in various positions. Knives lying flat were giveaways, but when viewed lengthwise, they had very little visible surface. Explosives were a whole different story. A plastic explosive like C4 is organic and dense. It appears as a heavy orange mass. Unfortunately, a block of cheddar cheese looks roughly the same.

As we started testing with a moving scanner, Gary warned me that too many false positives would be a big problem. A "hair-trigger" strategy would get me flunked. Images with guns took about one second to identify. Clear bags took roughly five seconds to double check for blade edges. It was cluttered bags—with their multihued oranges, blues, greens and grays jumbled together—that were the killers.

I wish that more of our passengers could see the system from the perspective of a screener. It is here, at the front lines, where the conundrum of airport security is in sharpest relief: the fear of missing even the smallest thing, versus the likelihood that you'll miss the big picture when you're focused on the small stuff.

Clearly, things needed to change. By the time of my arrival, the agency was focused almost entirely on finding prohibited items. Constant positive reinforcement on finding items like lighters had turned our checkpoint operations into an Easter-egg hunt. When we ran a test, putting dummy bomb components near lighters in bags at checkpoints, officers caught the lighters, not the bomb parts.

I wanted to reduce the amount of time that officers spent searching for low-risk objects, but politics intervened at every turn. Lighters were untouchable, having been banned by an act of Congress. And despite the radically reduced risk that knives and box cutters presented in the post-9/11 world, allowing them back on board was considered too emotionally charged for the American public.

We did succeed in getting some items (small scissors, ice skates) off the list of prohibited items. And we had explosives experts retrain the entire work force in terrorist tradecraft and bomb-making. Most important, Charlie Allen, the chief of intelligence for the Department of Homeland Security, tied the TSA into the wider world of U.S. intelligence, arranging for our leadership to participate in the daily counterterrorism video conference chaired from the White House. With a constant stream of live threat reporting to start each day, I was done with playing defense.

But the frustrations outweighed the progress. I had hoped to advance the idea of a Registered Traveler program, but the second that you create a population of travelers who are considered "trusted," that category of fliers moves to the top of al Qaeda's training list, whether they are old, young, white, Asian, military, civilian, male or female. The men who bombed the London Underground in July 2005 would all have been eligible for the Registered Traveler cards we were developing at the time. No realistic amount of prescreening can alleviate this threat when al Qaeda is working to recruit "clean" agents. TSA dropped the idea on my watch—though new versions of it continue to pop up.

Taking your shoes off for security is probably your least favorite part of flying these days. Mine, too. I came into office dead set on allowing people to keep their shoes on during screening. But, contrary to popular belief, it isn't just Richard Reid's failed shoe-bomb attempt in December 2001 that is responsible for the shoe rule. For years, the TSA has received intelligence on the terrorists' footwear-related innovations. Some very capable engineer on the other side is spending a lot of time improving shoe bombs, which can now be completely nonmetallic and concealed in a normal street shoe. There's still no quick way to detect them without an X-ray.

Airport security in America is broken. I should know. For 3½ years—from my confirmation in July 2005 to President Barack Obama's inauguration in January 2009—I served as the head of the Transportation Security Administration.

You know the TSA. We're the ones who make you take off your shoes before padding through a metal detector in your socks (hopefully without holes in them). We're the ones who make you throw out your water bottles. We're the ones who end up on the evening news when someone's grandma gets patted down or a child's toy gets confiscated as a security risk. If you're a frequent traveler, you probably hate us.

More than a decade after 9/11, it is a national embarrassment that our airport security system remains so hopelessly bureaucratic and disconnected from the people whom it is meant to protect. Preventing terrorist attacks on air travel demands flexibility and the constant reassessment of threats. It also demands strong public support, which the current system has plainly failed to achieve.

The crux of the problem, as I learned in my years at the helm, is our wrongheaded approach to risk. In attempting to eliminate all risk from flying, we have made air travel an unending nightmare for U.S. passengers and visitors from overseas, while at the same time creating a security system that is brittle where it needs to be supple.

I was initially against a ban on liquids as well, because I thought that, with proper briefing, TSA officers could stop al Qaeda's new liquid bombs. Unfortunately, al Qaeda's advancing skill with hydrogen-peroxide-based bombs made a total liquid ban necessary for a brief period and a restriction on the amount of liquid one could carry on a plane necessary thereafter.

Existing scanners could allow passengers to carry on any amount of liquid they want, so long as they put it in the gray bins. The scanners have yet to be used in this way because of concern for the large number of false alarms and delays that they could cause. When I left TSA in 2009, the plan was to designate "liquid lanes" where waits might be longer but passengers could board with snow globes, beauty products or booze. That plan is still sitting on someone's desk.

The hijackings of the 1960s gave us magnetometers, to keep guns off planes. After the Pan Am 103 bombing over Lockerbie, Scotland, a small amount of international checked baggage was scanned and people were required to fly with their luggage. After 9/11, the TSA was created and blades were banned.

Looking at the airport security system that we have today, each measure has a reason—and each one provides some security value. But taken together they tell the story of an agency that, while effective at stopping anticipated threats, is too reactive and always finds itself fighting the last war.

Airport security has to change. The relationship between the public and the TSA has become too poisonous to be sustained. And the way that we use TSA officers—as little more than human versions of our scanners—is a tremendous waste of well-trained, engaged brains that could be evaluating risk rather than looking for violations of the Standard Operating Procedure.

What would a better system look like? If politicians gave the TSA some political cover, the agency could institute the following changes before the start of the summer travel season:

1. No more banned items: Aside from obvious weapons capable of fast, multiple killings—such as guns, toxins and explosive devices—it is time to end the TSA's use of well-trained security officers as kindergarten teachers to millions of passengers a day. The list of banned items has created an "Easter-egg hunt" mentality at the TSA. Worse, banning certain items gives terrorists a complete list of what not to use in their next attack. Lighters are banned? The next attack will use an electric trigger.

2. Allow all liquids: Simple checkpoint signage, a small software update and some traffic management are all that stand between you and bringing all your liquids on every U.S. flight. Really.

3. Give TSA officers more flexibility and rewards for initiative, and hold them accountable: No security agency on earth has the experience and pattern-recognition skills of TSA officers. We need to leverage that ability. TSA officers should have more discretion to interact with passengers and to work in looser teams throughout airports. And TSA's leaders must be prepared to support initiative even when officers make mistakes. Currently, independence on the ground is more likely to lead to discipline than reward.

4. Eliminate baggage fees: Much of the pain at TSA checkpoints these days can be attributed to passengers overstuffing their carry-on luggage to avoid baggage fees. The airlines had their reasons for implementing these fees, but the result has been a checkpoint nightmare. Airlines might increase ticket prices slightly to compensate for the lost revenue, but the main impact would be that checkpoint screening for everybody will be faster and safer.

5. Randomize security: Predictability is deadly. Banned-item lists, rigid protocols—if terrorists know what to expect at the airport, they have a greater chance of evading our system.

In Richmond, Va., we tested a system that randomized the security procedures encountered by passengers (additional upper-torso pat-downs, a thorough bag search, a swab test of carry-ons, etc.), while not subjecting everyone to the full gamut. At other airports, we tried out a system called "Playbook," which gave airports a virtual encyclopedia of possible security actions and let local law-enforcement, airport and TSA officials choose a customized set of counterterror measures.

Implemented nationally, this approach would give to the system as a whole a value greater than the sum of its parts—making it much harder for terrorists to learn how to evade our security protocols.

To be effective, airport security needs to embrace flexibility and risk management—principles that it is difficult for both the bureaucracy and the public to accept. The public wants the airport experience to be predictable, hassle-free and airtight and for it to keep us 100% safe. But 100% safety is unattainable. Embracing a bit of risk could reduce the hassle of today's airport experience while making us safer at the same time.

Over the past 10 years, most Americans have had extensive personal experience with the TSA, and this familiarity has bred contempt. People often suggest that the U.S. should adopt the "Israeli method" of airport security—which relies on less screening of banned items and more interviewing of passengers. But Israeli citizens accept the continued existence of a common enemy that requires them to tolerate necessary inconveniences, and they know that terror plots are ongoing.

In America, any successful attack—no matter how small—is likely to lead to a series of public recriminations and witch hunts. But security is a series of trade-offs. We've made it through the 10 years after 9/11 without another attack, something that was not a given. But no security system can be maintained over the long term without public support and cooperation. If Americans are ready to embrace risk, it is time to strike a new balance.

—Mr. Hawley is the author of "Permanent Emergency: Inside the TSA and the Fight for the Future of American Security," to be published April 24 by Palgrave Macmill