DOT Secretary Duffy wants another $10 billion for ATC modernization, on top of the $12.5 billion approved in July 2025, aimed mostly at software, telecom, radios, radars, and related infrastructure.
But the real shortage is not in the systems and equipment. It is in controllers who are fully trained and working traffic.
That is the first point to keep in mind: The FAA is talking about system modernization and hires, but the shortage is in fully trained Certified Professional Controllers, or CPCs.
The Government Accountability Office, or GAO, numbers show why.
From 2017 through 2022, 106,533 people applied for air traffic control jobs. Only 2,258 made it to the far end, and even that group still included some people in training. That is only a 2.1% yield. The other 97.9% washed out along the way.
Now set that beside the FAA’s current workforce plan. The 2,200 figure for fiscal year 2026 is only one year in a larger hiring push. FAA laid out 2,000 hires for 2025, 2,200 for 2026, 2,300 for 2027, and 2,400 for 2028. That is 8,900 planned hires across four fiscal years.
Sounds strong. But it runs into the same problem GAO already showed in the 2017 through 2022 period. A big hiring push at the front end does not automatically turn into fully trained CPCs at the back end. That is the point people ought to be looking at. The GAO showed a long, thinning pipeline. The FAA is now trying to push a bigger multiyear hiring plan through that same kind of pipeline.
The FAA’s own workforce plan says the problem is not just need. It is also training capacity. Contract instructors are one limit. Academy capacity is another. Field facilities can handle only so many trainees without slowing or weakening training.
That bottleneck does not stop there. It follows trainees into the field. The FAA says new controllers spend several months at the Academy and then years more in training before they become CPCs. GAO says the full process can take two to six years. It is a long, leaky pipeline, and many do not make it through. That is why a hiring goal is not the same as a staffing fix. The finished product is not an applicant, not a hire, and not even a trainee. The finished product is a fully certified controller working live traffic without supervision.
> But the real shortage is not in the systems and equipment. It is in controllers who are fully trained and working traffic.
Have you read the FAA’s proposed plan regarding system and equipment upgrades? If so, do you believe the upgrades will actually reduce staffing needs?
> A big hiring push at the front end does not automatically turn into fully trained CPCs at the back end.
Do you know how many applications the FAA plans to accept? Couldn’t the FAA account for the washout rate when setting the number of applicants? Does your appraisal take into account improvements in training technologies that are part of the proposal?
Yikes! You suggest that the FAA could “account for the washout rate” when filling the pipeline at the front, but the numbers already say that they need around half a million applicants just to hit their four year plan. That is not remotely sustainable – to keep up with the long term mission they will eventually need to recruit pretty much the entire adult population of the USA. Solution – they need to fix the wash-out rate.
I believe new training systems and equipment is part of the upgrade plan. Have you read it?
“That bottleneck does not stop there.”
The bottle neck not only doesn’t stop there, it doesn’t start there.
The system is/has reached saturation. There’s only so many aircraft allowed on the runway, only so many aircraft you can have going to the same airport at the same time, only so many aircraft in a described airspace that a controller can operate.
Add another 10,000 controllers and you’ll still have the same bottlenecks.
And the equipment is not in shortage per se, but in a different state of bottleneck where a rats nest of ad-hoc solutions makes for a state of atrophy.
You are correct. But, at least with those extra controllers, current staffing issues could be eliminated and, among other things, eliminate forced overtime. It’s something.
Robert, I think you nailed the larger picture.
Staffing is one bottleneck, but it is not the only one. Runway capacity, airport arrival rates, saturated sectors, aging equipment, workaround procedures, and controller workload all stack on top of each other.
That is why I do not see modernization and staffing as separate issues.
They are tied together. Better systems may help, but they do not instantly create Certified Professional Controllers.
And more controllers help, but they do not magically create more concrete, more gates, or more arrival capacity.
The NAS is not short of just one thing. It is short of margin.
Yup, attack the washout rate to minimize wasting training on those who won’t make it and improve success rate. Simple answer to enormously complex problem: (1) examine characteristics of those who make it through vs those who don’t and feed that info into the recruiting process and (2) identify specific reasons people washout and attack those reasons. Determine what should have been done differently to help more of them succeed (customize the training to the candidate? improve mentoring? better pizza?) Rinse and repeat forever.
Agree with your assessment, Raf. As a retired ATCS (GSP, CVG, COS D01[Denver TRACON]) I also have observed that the challenge is the high washout rate. The nature of the job as it now exists requires a unique skillset that is not that commonly found. At its best it is still vulnerable to human error or any slight inattention or incapacity. I propose that the solution is to fundamentally change the way ATC is done, most likely transitioning to an AI based format, with human oversight.
How will we grow enough new, capable human overseers when most of the real work is done by AI? I’ve seen that problem (lack of new human experts) manifest itself over and over again whenever automation takes over most of the work from humans. It’s possible to make it work but only if you plan for it and work it into the new system.
>I propose that the solution is to fundamentally change the way ATC is done
I believe that is part of the plan from what I have seen. They are trying to streamline processes that exist. Catering to both those trained in the current system and new hires. The problem with a complete overhaul is what do you do with the people who are trained on the old system? how long will it take to get those people up to speed on the new process. I agree with utilizing AI to some extent will help with the streams of information.
Back when I was interviewing for jobs many (many) years ago, I used to enjoy asking HR “what happened to the previous guy”? Turned out that a lot of the open positions had “unreasonable levels of performance expectation”. The ATC environment as currently staffed and equipped is just too hard at today’s traffic levels. As noted above, there are unique individuals who can handle it, but they are rare. I should note in closing that a good controller is a huge confidence booster when the weather forecast goes down the tubes. Thank you ATC.
Raf maybe you solution is lower the standards! 😉
Jeffe, no. Lowering standards is exactly the wrong answer.
A good academy should turn out good results and good controllers. If the attrition is that high, then the problem is not solved by lowering the bar. It is solved by asking why the training system is losing so many people before they become useful Certified Professional Controllers.
That points to a failure to educate, screen, mentor, and train properly through the whole pipeline.
ATC is not a place for soft standards. But when the system washes out large numbers year after year, that is not something to shrug off as normal. It means the process needs a hard look.
Better selection, better instruction, better simulator time, better facility training, and better mentoring can improve the yield without weakening the standard.
Hiring numbers are not controller numbers. Academy seats are not Certified Professional Controllers. The finished product is a controller who can work traffic safely, unsupervised, under pressure.
That is the number that matters.
I don’t disagree with your reply BUT just about every aviation occupation has high attrition rates. For example the majority of people who start flight training never finish with a private pilot certificate. Same in the military aviation and special forces branches. Some things in life are really hard. Yesterday I flew to an airport that is a training tower. I was told to hold outside of the Delta because there were three aircraft in the pattern. Once I entered the pattern with one other aircraft in the pattern I was instructed to do two 360s as well as the RV in front of me because another aircraft reported a straight in to the runway. That trainee will likely never become a fully qualified controller.
And the “under pressure” is truth. Pressure is not continuously there. But when it is, there’s only one person who can salvage the mess, actually keep them from running together. Only you. And training must push the trainee to that level enough times that he/she can show they can do it, every time. Many controllers do not want to work to that level, an uncomfortable level, sometimes scary. Not many folks can do and either wash or bail out. So, we must reduce complexity. More runways, less aircraft, more staffing. Reduce it down to a level 3 activity, and so many more will succeed. Oh yea, back when those levels existed. ORD Tracon was as complex as any traffic picture I’d seen at the time. We would occasionally get good folks to bid in, good experience, and a desire to succeed. But when it came time to check out on ORD arrivals, the pressure to do it right, never wasting space or near loosing it, many of those folks just could not push it over the top. But because they were still a warm body, if they had checked out on departures and satellite positions, they were kept. They still provided much needed staffing relief. Another way to do it. It’s not really going to get better with staffing until complexity is reduced. My opinion. And don’t forget, for every new trainee, there must be a veteran CPC available and on duty to train that person. Some controllers are excellent trainers, enjoy doing it, and some just suck, So much to consider.
Hi roger,
exactly! i fly thru the big ‘O’ ord 2 x a yr on trips to airventure. I see now the runways are east/west 3 huge parallels to the north and 3 to the south. Looks like each rwy has a tower? Years ago the runways were in a triangle and the small little brown colored atc twr stll standing below the big green colored tower. Hats off to anyone that worked in that small tower and the old rwy config. And in the middle of winter on a snowstorm. Bravo! And ord approach for feeding the tower!
My FAA time was all sf bay area, challenging with 3 big airports and lots of busy smaller airports , and the topgraphy with lots of hills , mountains etc. The mva’s were a big challenge. Luckily good weather most of the time.
Thanks Michael. My time there was the old airport. One of our controllers described it as a bunch of runways that looked like they were designed like pick up sticks, and with a single tower and way understaffed Tracon. All the newer parallel runways should be making life a bit easier now, although I think they are once again number one in traffic activity. Someone said the new towers are seldom used at the same time because of staffing…..of course.
Hi Raf, all,
Thanks for the article and comments. As a 38 yr atc retired, 28 FAA and 10 USAF,
Many of us could say the same thing! Staff up! its been a disaster since 3 aug 1981.
People are the most important resource. Doesnt matter the latest equipment, its the staffing. The latest presser by admin Bedford in regards to TOC is another faa doozy.
Two big events have really tanked the hiring in the last 15 years. First was to get rid of the atc skills assesement test, and the BQ qustionaire, mainly to hire lower level IQ trainees. Then the Covid virus and academy shutdown. Hire the best and brightest,
trust me it was no fun trying to train the untrainable. We also have new generation
of youngsters who have grown up vastly different than the middle age boomers.
This job is not going to be handed to you. It will be the hardest thing you will ever accomplish, yet the most rewarding. It is not for the meek or average . I think direct
hires to facilities for ex military atc, and the cti students is a good idea. OTS off the street hires go to the academy. Need to get more people into the system faster.
The FAA HR hiring office needs help too with staffing. That was a huge bottleneck when i got hired in 1990. Trust me working 6 day workweeks 10 hour shifts, rotating shifts
is a long haul, and when i retired i was completely exhausted , mentally and physically.
Lots of new young pilots out there, i also think the FAA should look into more waivers
for working past age 56 if the employee wants to. On some atc forums i read there is lots of fustration with people at lower level facilities who have no career progression,
due to short staffing at their facility, and cannot be released . Some have quit because
they dont make enough to buy a house. They went on to other careers. The other problem has been filling supe slots at big tracons or centers with people from low level towers that could release them. They couldnt pick from internal bids due to low staffing .
How would you like it that your new supe has never worked radar , and is now your boss and is evaluating you? For gods sakes, Staff up the system, have career progression, hiring and promotions based on merit, instead of the latest bandaid boondoggle.
Agreed. People are the backbone of the system.
New equipment helps, but it cannot replace trained controllers, and supervisors who are knowledgeable about the work.
Staff up, hire based on merit, provide proper training.
Hi Raf,
Someday would be great to meet up. im usually in palm springs, are you in thermal?
im headed to my sf bay area home then to airventure,
back for new knees then back to ps this fall. i have a hangar at KUDD.
See you during the fall. 👍🏻
MIchael, totally agree. Also retired ATC, 31 years all FAA. Back when I hired in, 1977, terminal controllers who passed the Academy were assigned lower level towers, mostly VFR. It was expected one would gain experience then move up to more complex facilities with radar. The rate to FPL was about 40%. Same with sups, I was told you can’t be a supe in the facility where you were an FPL (CPC), have to transfer. Well, Congress’ inconsistent budgets and continuing resolutions led to Academy shutdowns, lack of move money, and eventually contract towers, so the lower level pipeline was shut off. The answer from the higher ups was recycle the trainee failures, yes you could be a supe where you were an FPL, and you may have been accepted at another facility but your current facility is too low on staffing and can’t release you, maybe for over a year! Overtime was always considered as price concious for the agency compared to another body with benefits. Never mind what that does to the health of the controllers working six days a week!
I know a contractor at the Academy and they claim the incoming hires are not appropriate for the job. Can’t get there on time, have hearing aids, have Tourett’s syndrome, have childcare problems, the list goes on. The job is not easy and you have to push yourself. High IQ doesn’t always lead to success, it’s more a function of can you look at a two-dimensional presentation and convert it to three-dimensional in your head. The training tools are much advanced from when I was first hired, so don’t blame that. I would have loved to train on the tower simulators! It’s all about the right hires and the willingness to staff to the levels needed for proper leave and internal movement.
Hi DSloss,
Ive heard the same about the trainees. incredible.
So it shows they are still not hiring the best and brightest. Disabled, hearing aids, tourettes,
etc, how could they pass a faa class 2 medical.? Box checkers .
people quitting , say its too hard, blame everyone else but themselves. A sad state of affairs.
If you cant take the pressure of the academy, you will NEVER make it out in the facilities.
We are in trouble . The younger generation to pull from, 87% ineligible to enlist in the military.
So how could they even pass the academy? People at a young age with too much baggage.
I would pull from previous atc military controllers, send direct hires. Dedicated and disiplined.
Ready to roll.
Im still trying to process bedfords TOC theory of constraints. nonsense. .
Another faa boondoggle. Remember the hiring of a lady , for big $$$, who claimed to be the lost spirit of atlantis, was holding ‘training ‘ sessions in seattle for faa supervisors. Literally roped together,
men and women who were forced to go use the restroom together! lol i guess this was
‘team’ bonding. Incredible! All of us retired have seen the craziest nonsense over the years.
if you need 3000 the next 3 yrs just to keep even, you better be hiring 9000 because 2/3 wont make it.
We banter around $12.5 billion dollars as its pocket change, with Duffy announcing a no bid contract award to a surveillance company for hardware/software/radar “upgrades” that has it’s beginnings as a company in 2017. It’s job is to bring ATC into the modern era by mixing the equivalent to a Radio Shack/Tandy 486 computer complete with it’s latest floppy disc mated to an iPhone 17 with MicroSoft’s latest version of Apple’s most recent IOS update requiring a purchase of Office 360, Windows 10, and Norton’s antivirus package. Then we attempt to teach people how to integrate landline conversations which would include the knowledge of rotary phone operation, pencil written notes on slides, looking outside of the cab with binoculars, while using Alexa/Siri female voice activation, often using two separate frequencies for communications with inbound aircraft at the same time (of which inbound aircraft cannot hear both frequencies at the same time ala DC helo/RJ midair), often forcing one person to handle clearance delivery, ground control, and tower operations at the same time…coordinating with Center using an equally loathsome mish-mash of 40+/30+/20+10 year old technology mix (remember NexGen?)…10 hours a day 6-7 days a week…and can’t figure out a 98% washout rate?????…SERIOUSLY?
First we have to have a well engineered, USA manufactured(certainly NOT Boeing, Lockheed-Martin), USA sourced and backed ATC hardware/software/parts/maintenance package of nationwide , standardized equipment, including radar/weather radar integrated with reliable/safe/secure internet capability that does not rely on Elon’s Starlink, a hodge podge ground system, Sirius XM,etc… competitively bid out between several USA based and operated companies, and evaluated by aviation knowledgeable, CPC level, experienced FAA/ATC personnel, test flown by qualified GA/Airline/Corporate/FAA pilots in real airplanes.
Or..we can continue to do what we have been doing for the last 75+ years and expect a different outcome?
Standardized Training, meeting staffing requirements, and hardware/software system wide improvements cannot be done isolated from each other. Nor can we let our largely aviation ignorant legislators decide what we spend, who is going to create and build the next version of ATC “modernization”, and decide the proper course of training action to meet current and future, professional ATC staffing requirements. Right now, we have the blind leading the blind, spending money we don’t have, on systems that currently do not exist, but are attached to a letter of intent, from a surveillance company that is less than 10 years old, already in process of spending $12.5 billion dollars fresh off the printing press, and need another $11 bil. What could possibly go wrong?
History 101, I agree with the core point.
Modernization is not just buying equipment. Now add BNATCS, the Brand New Air Traffic Control System, and the 2028 timeline gets even harder to understand.
New systems bring downtime, testing, certification, maintenance problems, controller retraining, and a learning curve. Meanwhile, the old system still has to run safely every day.
So the FAA is trying to modernize the system, fix staffing, train new controllers, retrain current controllers, and keep traffic moving at the same time.
That is a lot more complicated than a press release makes it sound .
What could possibly go wrong?