1
Introduction
As commercial airline traffic rebounded in 2022 following the worst effects of the COVID-19 pandemic, so did flight cancellations and delays that disrupted travel for more than 10 million passengers (USDOT 2022). Amid peak holiday travel at the end of 2022, heavy winter storms contributed to cancellations of more than 32,000 flights, affecting millions of passengers. Throughout 2022, the U.S. Department of Transportation (USDOT) had responded to the increase in cancellations and delays with scrutiny of airline scheduling practices and staffing and increased pressure on airlines to meet their obligations to passengers (GAO 2023). However, the airline association representing most major carriers responded by attributing some of the 2022 delays and cancellations, including on days with no weather disruptions, to shortages in air traffic control (ATC) staffing, particularly at key facilities in the New York-New Jersey metropolitan area and Florida (A4A 2022).
By early 2023, as traffic demand continued to grow, flight disruptions did as well. Even though the rates of close calls in the air and on runways were similar to prepandemic years, several occurred within a short period, generating a safety alert by the Federal Aviation Administration (FAA) (Holpuch 2023). By June 2023, the USDOT Office of Inspector General (OIG) reported that many of FAA’s critical ATC facilities across the country were facing significant shortages of controllers and that the agency lacked a plan to address them (USDOT OIG 2023).
For decades, FAA had been using staffing models for planning the controller workforce at its ATC facilities, which the committee refers to as the
ATC workforce.1 The ATC workforce includes the Certified Professional Controllers (CPCs) who account for most of the work controlling traffic and who have demonstrated mastery of each of the primary tasks (referred to as positions) within their respective facilities. FAA’s facility staffing models are structured to estimate staffing levels of Position-Qualified Controllers (PQCs) needed to meet the demand for managing traffic on a peak (90th) percentile travel day. The model output is commonly referred to as facility staffing standards or Staffing Standard Targets, which are provided in an annual publicly available Controller Workforce Plan (CWP).
In 2012, FAA established a Collaborative Resource Working Group (CRWG) to develop models and tools to allocate new hires and transfers to facilities that most need them. The members were made up of staff from FAA’s Air Traffic Organization (ATO), Office of Finance and Management (initially), and the National Air Traffic Control Association. For reasons that are unclear, by 2016 the CRWG had developed a separate modeling approach for estimating the number of CPCs needed to manage traffic at facilities. The CRWG targets for Fiscal Year (FY) 2023–2024 would increase the number of CPCs by roughly 30% (although it would take many years to hire and train sufficient controllers to CPC status).
In the FAA Reauthorization Act of 2024, enacted on May 16, 2024, Congress tasked the National Academies of Sciences, Engineering, and Medicine (NASEM) with convening an expert committee to reexamine FAA’s ATC staffing models and plans as NASEM had done in three previous reports, most recently in 2014. Anticipating this legislative request, FAA had already contracted with NASEM several months earlier for the same purpose. The committee’s Statement of Task (SOT) (provided in brief directly below and completely in Box 1-1), approved by NASEM and FAA, asked the committee to
- assess FAA’s progress in responding to the NASEM 2014 report recommendations to improve its staffing models and execution of its staffing plans;
- identify and assess factors relevant to controller staffing levels and planning by analyzing historical data and metrics;
- compare the staffing models and approaches FAA had been using for decades with the CRWG approach;
- consider how FAA controller hiring and training plans and their execution affect current controller staffing levels; and
___________________
1 FAA often uses the Air Traffic Control Specialist workforce, but it can include supervisors. This report focuses on facility frontline air traffic controllers and trainees.
- identify the future needs of the ATC system and their potential impacts on controller staffing requirements and planning, including developments such as the integration of new users, technologies, and procedures.
After the 2024 FAA reauthorization legislation passed, FAA amended the contract with NASEM to elaborate on and ensure that the committee addressed specific items in the enacted legislation regarding aspects of the models that should be compared, the budgetary impacts of adopting the CRWG method, and potential future changes in aviation that could affect the number of controllers needed (see Box 1-2).
The committee’s report, herein, addresses all of the above items organized in a manner described at the end of this chapter. The next section provides high-level background on FAA’s ATC system and controller staffing at the agency’s 313 ATC facilities. A subsequent section provides an overview of the various models and the roles they play in the actual staffing of these facilities. Before concluding with an overview of the report’s organization, a section on how facilities are actually staffed provides an important conclusion from this introductory discussion, the significance of which carries through the rest of the report.
BACKGROUND
FAA’s ATC workforce of roughly 13,460 Full-Time Equivalent controllers as of the end of FY 2024 plays critical roles in the safety and efficiency of the U.S. aviation system. FAA’s management of the airspace is delegated to ATO. ATO’s controllers provide guidance to roughly 45,000 daily flights serving nearly 3 million passengers and cargo operations in a 24/7/365 operation. The workforce’s primary mission is safety, achieved by providing for the safe separation between aircraft in the air and on the ground at airports serving commercial, military, and general aviation aircraft and sharing weather and traffic advisories with pilots. In addition, FAA’s staffing models and annual CWPs, and ATO’s execution of those plans, strive to provide a sufficiently large staff of controllers at each of its 313 facilities to allow for the efficient movement of traffic every day of the year. However, safety is paramount. When ATO is short-staffed relative to traffic demand, it limits the number of flights in the air to what its workforce can handle safely.
In addition to CPCs, each ATC facility is staffed with new hires called Developmental Controllers, whose primary task is to complete training and become certified on each of the specific positions within their respective facilities (or areas within the larger facilities). These new staff will ultimately
BOX 1-1
Committee Statement of Task (SOT)
An ad hoc committee will conduct a study to inform the Federal Aviation Administration’s (FAA’s) interest in developing objective, science-based approaches for setting future air traffic controller staffing targets to ensure the safe and efficient operation of the national airspace system. The committee will compare and evaluate competing controller staffing models, examine the array of factors that contribute to current staffing levels, and assess future needs of the air traffic control system and their potential impacts on staffing requirements.
Currently there are two models for developing FAA’s future air traffic controller staffing targets; one developed by FAA’s Office of Financial and Labor Analysis and the other developed by an FAA collaborative resource working group with involvement by the National Air Traffic Controllers Association. The committee will review the structure, assumptions, data sources, and other components of the two models, consider why they have produced different aggregate and facility-level staffing targets, and assess their respective strengths and weaknesses.
The committee will make recommendations to FAA and Congress, as appropriate, to improve methods for planning the future controller workforce taking into account experiences with regard to factors such as controller task loads; performance efficiency; time spent on nontraffic duties; training requirements; hiring, retention, transfer, and retirement patterns; leave usage; and scheduling needs and challenges. In making recommendations, the committee will also consider how developments such as the integration of new users and technologies and new navigation and traffic control procedures may impact future controller staffing requirements and planning. The committee will also review the progress made by FAA in furthering the recommendations of previous National Academies of Sciences, Engineering, and Medicine reports on controller staffing approaches.
To fulfill the SOT, the study committee will:
- assess FAA’s progress in furthering the recommendations in (the NASEM 2014 report) The Federal Aviation Administration’s Approach for Determining Future Air Traffic Controller Staffing Needs;
- identify and assess factors relevant to controller staffing levels and planning by analyzing historical data and metrics for factors such as operations workload, Time on Position (TOP), Other Duties, planning decisions and training execution, certification progression, overtime and leave usage, and planning and scheduling activities;
- compare the FAA and Collaborative Resource Working Group (CRWG) staffing models and methodologies by reviewing similarities and differences in their structures, assumptions, data sources, and other components; their treatment of factors relevant to controller staffing requirements; the targets they produce for a varied selection of ATC facilities; and how well they conform to industry best practices for workforce planning;
- consider how FAA controller hiring and training plans and their execution affect current controller staffing levels; and
- identify future needs of the ATC system and their potential impacts on controller staffing requirements and planning, including developments such as the integration of new users, technologies, and procedures.
BOX 1-2
Additions to the Contract Following Passage of the FAA Reauthorization Act of 2024
Compare the FAA and CRWG Staffing models to determine Certified Professional Controller (CPC) operational staffing targets necessary to meet facility operational, statutory, contractual and safety requirements, including
- the availability factor multiplier and other formula components;
- the independent facility staffing targets of CPCs able to control traffic;
- air traffic controller position utilization;
- attrition rates at each ATC facility operated by the administration; and
- the time needed to meet facility operational, statutory, and contractual requirements, relevant resources to develop, evaluate, and implement processes and initiatives affecting the national airspace system.
Per the 2024 FAA Reauthorization, FAA is directed to “determine which staffing models and methodologies evaluated pursuant to this subsection best accounts for the operational staffing needs of the air traffic control system and provide a justification for such determination.” We recognize that neither model is perfect, therefore FAA seeks recommendations for a path forward for an improved methodology or process for selecting the target staffing numbers.
Examine the current and estimated budgets of FAA to implement the FAA controller staffing standards included in the 2023 Controller Workforce Plan (CWP) in comparison to the funding needed to implement the CRWG CPC operational staffing targets.
Identify future needs of the Air Traffic Control System and their potential impacts on controller staffing requirements and planning, including projected air traffic in the airspace of each ATC facility operated by the administration.
replace the CPCs or other controllers who leave the facility due to retirement, transfer, promotion, training failure, or other reasons. However, it takes 2–3 years or more of training to reach CPC status, depending on the type of facility and the complexity of the traffic managed, as described in Chapter 6. In addition, when a CPC transfers from one facility to another, they become a CPC-In Training (CPC-IT) until reaching full certification at the new facility. This recertification process can take 1.5 years or more.
For largely externally imposed reasons explained in Chapter 2, over the period between FY 2013 to 2023, ATO experienced an average of 1,330 losses annually from its ATC workforce due to attrition but was only able to hire an average of 1,210. The total number of CPCs declined by about 950. As also explained in Chapter 2, losses were particularly acute in key facilities that serve a high proportion of daily commercial traffic.
Moreover, because CPCs provide the required On-the-Job Training (OJT) to Developmental Controllers and CPC-ITs, the 11-year decline in CPCs created a potential bottleneck in the training pipeline. With fewer CPCs available to provide OJT to trainees, the certification process is potentially lengthened or will require overtime to maintain training schedules.
Although ATO began increasing its hiring targets in FY 2023 and 2024, current FAA’s Office of Finance and Management (AFN) projections at the time of hiring, attrition, and traffic indicated the organization would achieve adequate staffing in FY 2027–2028 (Safety Review Team 2023, Figure 4). The CWRG model’s recommended CPC targets, which call for higher staffing levels at facilities, would not be met until after 2032. Even if the total model-estimated total staff numbers are eventually reached across the overall workforce, this does not guarantee adequate staffing at each facility. Chapter 2 includes tables showing that about 60% of facilities are either below or above their model-estimated targets when expressed as +/–10% of the target.
Declining staffing at many facilities has led to increased reliance on overtime, raising safety concerns that prompted the FAA Administrator to commission an independent panel of sleep and fatigue experts. The panel’s April 2024 report recommended, among other things, that FAA provide more opportunities for rest and restorative sleep between shifts, especially those that span midnight (Rosekind et al. 2024), echoing similar recommendations from the NASEM 2014 report (NRC 2014). Coincident with the release of the 2024 report, the FAA Administrator issued an order that required longer rest periods between shifts and initiated elimination of the most fatigue-inducing shifts (see Chapter 3).2 However, this order cannot be fully implemented until revised schedules are developed in FY 2025, as scheduling is only accomplished once per year. Elimination of the most fatigue-inducing schedules is expected to occur in FY 2026.
Congress has also taken steps to accelerate the rebuilding of FAA’s workforce through the FAA Reauthorization Act of 2024,3 which included requirements that FAA (a) revise its screening of new hires (Sec. 417) to reduce the rate of training failures and (b) expand its hiring and training capacity to accelerate the growth in total headcount and the future number of CPCs (Secs. 415, 416, 437, and 439). Congress also tasked NASEM, through the committee for this report, to provide guidance on whether FAA should adopt the CRWG model to estimate the number of needed CPCs or continue to rely on the models it has traditionally used. In addition to assessing whether FAA should adopt the CRWG model’s higher targets,
___________________
2 See https://www.faa.gov/newsroom/statement-faa-administrator-mike-whitaker.
3 Public Law No. 118-63 (05/13/2024). See https://www.congress.gov/bill/118th-congress/house-bill/3935.
Congress also asked for an evaluation of the budgetary impact of hiring to the CRWG’s targets.
FAA’S OVERALL MODELING APPROACH
Congress’s request for this report emphasizes the models used to determine the staffing levels appropriate at each facility. These facility staffing models estimate the number of PQCs needed to ensure that each facility has enough staff available to come to work for the 90th percentile day of traffic to cover all the positions that must be staffed to safely separate the expected traffic. However, FAA also employs other models in addition to those used to estimate adequate facility staffing levels. This section provides an overview of FAA’s overall modeling process to clarify the role of the facility staffing models cited in the study charge relative to the setting of an annual hiring goal to ensure that enough trainees are in the pipeline to fill current and future staffing gaps at facilities.
FAA describes its overall modeling approach using a construct to depict the role of facility staffing models in the overall process of providing each facility with a controller workforce adequate to meet each facility’s traffic demand (see Figure 1-1). The models for estimating the staffing standards are maintained by AFN. Both the AFN and CRWG facility staffing models provide an estimate for the “A” in Figure 1-1, which is the target level of staffing required at each facility to meet traffic demand (i.e., the staffing standard [or target] for each facility).4 The “B” quantity represents the sum of the staffing levels based on the CPCs and CPC-ITs at each facility at the time the models are run each year. The “C” quantity accounts for changes in the workforce and is where AFN’s attrition models come into play for the existing workforce. (The CRWG models do not have a separate set of attrition models because to date the CRWG facility estimates have only been used in prioritizing where new hires are placed for training and to oversee the transfer process [see Chapter 6].) “D” is the gap between where staffing is at the times models are run and where the facility staffing models indicate it should be after accounting for attrition, referred to in this report as the “staffing gap.” The staffing gap is a primary input into
___________________
4 The staffing standard target developed by FAA’s traditional models represents CPCs as well as fully certified controllers who transfer across facilities and who must train to recertify (CPC-1T), a process that normally takes 1.5 years or more. The facility staffing models for “A” do not estimate how many other trainees a facility needs to replace fully certified controllers who transfer from a facility or otherwise leave the workforce; in the AFN modeling process the need for trainees is estimated from loss of controllers to attrition in step “C.” The target developed by the CRWG method is for fully certified controllers only. The differing targets and their importance are described in Chapter 4.
SOURCE: FAA figure provided to the committee with explanatory text developed by the committee.
the hiring goal set each year. As described in Chapter 2, ATO also takes into account budgetary and other considerations when setting the annual hiring goal. The comparison made between the two modeling approaches in Chapter 4 focuses on the different assumptions and methods they use in estimating “A.”
PLAN EXECUTION
The difference between what FAA’s staffing models predict is needed at facilities and how they are actually staffed is an important and often unappreciated distinction in the debates and reports about FAA’s facility staffing levels. As described in the next chapter, due partly to the hiring constraints FAA experienced between FY 2013 and 2023, most ATC facilities are not staffed within +/–10% of the targets determined by FAA’s traditional staffing models. Understaffing at key facilities has contributed to the traffic delays described above. However, because many facilities are not staffed according to what the models predict, actual facility staffing levels are attributable to the execution of FAA’s staffing plan, including hiring based on modeled estimates, not necessarily because the model predictions of the staffing needed at facilities are flawed. As described in Chapter 3, it is difficult to validate whether the models ATO has traditionally relied on are predicting the staffing levels that are adequate for facilities, but when ATO is not staffing each facility at the levels the models predict, then both model
predictions and execution of the staffing plan have to be taken into account in assessing the adequacy of facility staffing.
How staff are deployed to facilities where they are needed is referred to as “plan execution,” as described further and analyzed in Chapter 6. Plan execution is the process by which candidates are recruited, screened, and hired. Candidates are trained initially at the FAA Academy, for those hires without previous controller experience, and then placed in the field for a few years of OJT until they reach full CPC status. Thus, although the staffing models estimate the number of staff each facility should have, the hires that FAA makes each year serve what FAA believes it needs in total across all facilities. New hires are not specially designated for individual facilities that need them even though the hiring goal is based on the sum of modeled estimates at each facility. Staffing of facilities according to what the models estimate depends on the number of hires, the placement of hires, transfers from within the existing workforce from facilities that have surplus staff to those that need them, and the attrition each facility experiences. Given individual controller preferences for workplace location, willingness of controllers to transfer, and uncertainty of training outcomes following transfer, staffing each facility according to modeled estimates has proven to be difficult.
Conclusion 1-1: The staffing levels at ATO’s individual ATC facilities are the result of facility staffing model estimates of needed CPCs and CPC-ITs, how many candidates FAA hires relative to the sum of these modeled estimates, how efficiently new controllers can be trained and certified, and how well ATO ensures that facilities are actually staffed according to modeled estimates. Regardless of which modeling approach FAA ultimately adopts for estimating appropriate facility staffing levels, the staffing levels estimated by the models will not be achieved unless hiring, training, and other related policies are aligned with modeled estimates.
REPORT ORGANIZATION
This report is organized around the specific elements of the committee’s charge. A brief discussion of the committee’s main findings and its main conclusion and recommendations appear in the Summary. A more complete presentation of findings, conclusions, and recommendations appears at the end of each chapter.
Chapter 2 provides additional background on FAA’s operation of ATC services. It begins with an explanation of how staffing is intended to ensure the safe separation of traffic and minimize fatigue on its workforce inherent
in shifts that cover facilities open 24 hours per day, 7 days per week, every day of the year. It describes the different types of ATO facilities, how staffing levels vary relative to what its models project are needed across facilities and the impacts that staffing shortages have on ATO’s ability to meet traffic demand. In response to specific items in the committee’s charge (see Box 1-1) to “examine the array of factors that contribute to current staffing levels” and “assess factors relevant to controller staffing levels and planning by analyzing historical data and metrics,” the committee also explains why the 12-year decline in CPCs occurred as FAA responded to external, and occasionally internal, constraints on its ability to hire enough staff to keep up with attrition and meet facility staffing standards.
Chapter 3 examines in greater depth issues with fatigue and safety associated with shifts that rotate throughout the week to cover morning and evening peaks in demand as well as midnight shifts. ATO’s past allowance of fatigue-inducing shifts interacts with the effects on fatigue of short-staffed facilities and makes their separate effects difficult to isolate. The NASEM 2014 report (NRC 2014) expressed concerns about fatigue-inducing shift patterns and implied that FAA should reduce its reliance on them and adopt science-based Fatigue Risk Management (FRM) practices. The chapter also discusses issues with validating the accuracy of facility staffing models using measures of workforce strain such as overtime hours worked that can contribute to controller fatigue.
Chapter 4 takes up the charge to compare the competing models that FAA has developed to estimate the number of controllers needed at each facility (the “A” part of the A−B+C=D modeling process). The chapter also responds to the item in the charge for assessment of whether FAA has “objective, science-based approaches for setting future air traffic controller staffing targets to ensure the safe and efficient operation of the national airspace system.” The committee’s comparison of the models addresses specific items requested by Congress in legislation that passed in 2024 after the project was already under way (see Box 1-2).
Chapter 5 goes into greater depth in explaining the A, B, C, and D steps in FAA’s overall modeling process for estimating the number of new staff that should be hired each year. It specifically addresses Congress’s stated interest in “attrition rates at each air traffic control facility operated by the administration” (see Box 1-2) and how the strength of the workforce at facilities should be evaluated.
Chapter 6 reviews FAA’s execution of its workforce plans. Whether the model-generated targets of facilities are met depends critically on recruiting, screening, hiring, placing, and training hundreds, if not thousands, of new staff each year. This chapter responds to the SOT item requesting the
committee’s assessment of “how FAA controller hiring and training plans and their execution affect current controller staffing levels.”
Chapter 7 responds to the SOT and congressional interest in assessing whether FAA’s workforce planning is preparing for future demands on the ATC workforce such as the potential demands of new generations of advanced autonomous or semiautonomous drones and small passenger aircraft, examples of which are already seeking design and operational certification by FAA.
Chapter 8 addresses congressional interest in “the current and estimated budgets of the FAA to implement the FAA Controller Staffing Standards included in the 2023 Controller Workforce Plan in comparison to the funding needed to implement the CRWG CPC operational staffing targets.”
In each of the preceding chapters, the committee addresses how FAA has responded to the relevant recommendations made by the committee for the 2014 NASEM report.
Biographical information regarding the expertise of the committee members responsible for this report is provided in Appendix A. Appendix B describes a database including all controllers on the FAA payroll since FY 2010 that the committee used to generate many of the tables and figures included in the chapters. Appendix C provides a listing of FAA’s 313 facilities, their staffing levels at the end of FY 2023, and the CRWG and AFN targets for each facility. Appendix D describes the derivation of the committee’s estimated cost of staffing-related delays to passengers and airlines that are referenced in Chapter 8.
REFERENCES
A4A (Airlines for America). 2022. “Letter from Airlines of America to Transportation Secretary Buttigieg.” Airlines, June 24. https://www.airlines.org/news/letter-from-airlines-for-america-to-transportation-secretary-buttigieg.
GAO (Government Accountability Office). 2023. Airline Passenger Protections: Observations on Flight Delays and Cancellations, and DOT’s Efforts to Address Them. https://www.gao.gov/assets/gao-23-105524.pdf.
Holpuch, A. 2023. “F.A.A. Issues Safety Alert After Runway Near Misses.” New York Times, March 22. https://www.nytimes.com/2023/03/22/business/faa-airlines-near-misses.html.
NRC (National Research Council). 2014. The Federal Aviation Administration’s Approach for Determining Future Air Traffic Controller Staffing Needs. The National Academies Press. https://doi.org/10.17226/18824.
Rosekind, M., E. Flynn-Evans, and C. Czeisler. 2024. “Assessing Fatigue Risk in FAA Air Traffic Operations: Report by Scientific Expert Panel on Air Traffic Controller Safety, Work Hours, and Health.” https://www.faa.gov/newsroom/media/Fatigue_Report.pdf.
Safety Review Team. 2023. “National Airspace System Safety Review Team: Discussion and Recommendations to Address Risk in the National Airspace System.” Federal Aviation Administration, November. https://www.faa.gov/NAS_safety_review_team_report.pdf.
USDOT (U.S. Department of Transportation). 2022. “Air Travel Consumer Report: December 2022, Full Year 2022 Numbers.” https://www.transportation.gov/briefing-room/air-travel-consumer-report-december-2022-full-year-2022-numbers.
USDOT OIG (Office of Inspector General). 2023. “FAA Faces Air Traffic Control Staffing Challenges as Air Traffic Operations return to Pre-Pandemic Levels at Critical Facilities.” https://www.oig.dot.gov/sites/default/files/FAA%20Controller%20Staffing%20and%20Training%20at%20Critical%20Facilities%20Final%20Report-06-21-23.pdf.
