Summary1
At the direction of Congress, the Federal Aviation Administration (FAA) tasked the National Academies of Sciences, Engineering, and Medicine (NASEM) with forming an independent and balanced committee of experts to study important concerns regarding shortages of staffing at air traffic control facilities and their causes. The most consequential elements of the study charge are to:
- examine the array of factors affecting current staffing levels of fully Certified Professional Controllers (CPCs) at many of FAA’s 313 facilities;
- compare facility staffing models that FAA has long relied on with an approach FAA developed collaboratively with the National Air Traffic Controllers Association (NATCA), the use of which would significantly increase the facility staffing targets for the controller workforce; and
- assess FAA’s adoption of recommendations regarding controller fatigue and safety made in a NASEM 2014 report on controller staffing.
See Chapter 1 for the full statement of task negotiated with FAA and for additional elements added by Congress that are addressed in the report.
___________________
1 This Summary does not include references. Citations to support the text and recommendations herein are provided in the individual chapters of the report.
The main findings, conclusions, and recommendations of the committee are summarized below and discussed more fully in individual chapters.
The committee’s overarching observation is that proposals to replace the facility staffing modeling approach that FAA has traditionally used with the collaborative approach developed with NATCA are misplaced for two reasons: (1) FAA hired only about two-thirds of what its traditional modeling approach called for in the Fiscal Year (FY) 2013–2023 period, partly due to a series of externally imposed hiring constraints, and (2) the majority of FAA’s facilities are not staffed to the targets (or standards) for which the traditional modeling approach has called. Since at least FY 2010, about 30% of FAA facilities have been staffed at levels that are more than 10% below their staffing targets as estimated by the traditional approach and about 30% of facilities at levels more than 10% above their targets. Moreover, the increase in overtime use experienced after FY 2013 attributed to staffing shortages is not limited to facilities staffed at levels more than 10% below their staffing standards; such widespread overtime use may be partly due to inefficient scheduling of the controllers available at facilities.
Conclusion S-1: The shortage of staff experienced at 30% of facilities due to insufficient hiring over a decade has been primarily due to (a) mostly externally imposed hiring constraints and (b) a misallocated workforce that includes controllers at overstaffed facilities, more of whom could transfer to understaffed facilities. Both causes of shortages at understaffed facilities are possibly exacerbated by inefficient scheduling. The shortage of staff experienced does not necessarily mean FAA’s traditional modeling approach is flawed and needs to be replaced.
FACTORS AFFECTING FACILITY STAFFING LEVELS
FAA experienced a series of externally imposed constraints on hiring in FY 2013–2023 that included two government shutdowns over budget and fiscal policy and the COVID-19 pandemic. Although these constraints were outside of FAA’s control, FAA’s annual 10-year hiring plans during this period did not convey an intention to reverse staffing losses it was experiencing until its FY 2023–2024 workforce plans significantly increased projected hiring. (See Chapter 2.)
Primarily as a result of hiring fewer controllers than those lost to attrition, 19 of the largest facilities have fallen 15% below their staffing targets as estimated with the traditional approach. These large, understaffed facilities serving the 30 largest airports account for about 6% of facilities, 27% of commercial operations, 40% of all delays, and 45% of other delays, which include those that are staffing-related. (See Chapters 2 and 8.)
Although FAA has been increasing hiring as directed by Congress in FY 2024, it can only hire as many new staff as it can train expeditiously due to constraints on training capacity at the FAA Academy and individual facilities. The principal constraint has been the shortage of former FAA controllers willing to serve as classroom trainers at the Academy and facilities. Through its National Training Initiative, FAA is working diligently to address this constraint. FAA is currently confident it has addressed this issue at the Academy. Attracting former controllers to serve as trainers at individual facilities is a long-standing problem in specific locations, typically those with high costs of living. Expanded and improved training is needed to rectify declining training success rates and increasing time required for new hires to fully certify at the largest facilities. Without further improvements in hiring and training, it would take an average of at least 4.8 years from tentative hire offers to reach CPC status in En Route Centers. For CPCs required at the largest Terminals, the bulk of which require transfers from a smaller facility, it would take an average of 5.5 years for new hires to reach CPC status. (See Chapter 6.)
FAA experienced roughly 700 voluntary transfers each year from FY 2010 to 2024, a number second only to the 1,200 average number of annual hires in contributing to adequate facility staffing. Since FY 2016, of an average 218 annual transfers from facilities 10% or more above their staffing targets, only 45 (20%) have gone to facilities more than 10% below theirs. There are at least 1,700 Full-Time Equivalents (FTEs) trained to work at Terminals in facilities more than 15% above their staffing targets. Some of these FTEs could transfer to the understaffed higher-level Terminals and recertify much more quickly (typically 1.5 years) than relying on new hires. (There are no En Route Centers that control aircraft between airports that are more than 10% above their staffing standards and therefore no surplus of En Route Center CPCs to transfer to En Route Centers that need them.) The process in place since FY 2016 to induce voluntary transfers from overstaffed to understaffed facilities has been only marginally successful. (See Chapter 6.)
Increased staffing for understaffed facilities will be facilitated by FAA’s continuing efforts to improve screening of applicants, increase hiring, and reduce training failures and time required to reach full certification.
Recommendation S-1:2 Even as it improves its models’ estimates as recommended below, FAA should make a concerted effort to staff individual facilities based on what its models estimate, particularly those that have an outsized impact on delays in commercial aviation. As part
___________________
2 S-1 stands for overarching recommendation 1. The numbers in parentheses following each recommendation represent the recommendations made in individual chapters.
of this process, FAA should use increased incentives to induce more transfers of CPCs from overstaffed to understaffed facilities. (R6-5 through R6-8)
Recommendation S-2: Congress should provide funding for (a) needed hiring to reach modeled staffing standards based on adjustments recommended below and (b) additional incentives for transfers from overstaffed to understaffed facilities. (R5-2 and R6-8)
In addition to widespread use of overtime, there are other perplexing trends across the controller workforce since FY 2010, such as the 4% fewer regular hours worked per FTE and 4% reduced Time on Position (TOP) controlling traffic per FTE, despite a 4% increase in operations and a 13% decline in overall FTEs. (See Chapter 3.)
Recommendation S-3: FAA should determine the causes of the perplexing trends in overtime, regular hours worked, and TOP and rectify them as needed. (R3-3)
COMPARISON OF MODELING METHODS
Both of the facility staffing modeling approaches the committee was asked to evaluate have strengths and weaknesses. The main strength of the traditional approach is its ability to forecast staffing needed to support future operational demand. FAA must be able to hire at least 3 years ahead of future traffic to allow adequate time for training to replace the annual loss of 9% of controllers who will retire or otherwise leave the workforce. However, the traditional approach may underestimate the time required for training and other essential duties (other duties) when controllers are not on position. In addition, the task load models (estimates of time required to control aircraft and communicate with pilots) used in the traditional approach have not been updated for at least a decade. Thus, they may not be accurately accounting for subsequent changes in technologies and traffic flow management that could affect the number of controllers required. The main strength of the collaborative approach is its reliance on a survey of facility managers as its starting point. However, a key assumption made in the collaborative approach most likely results in overestimates of the time needed for other duties and the number of controllers required at individual facilities. (See Chapter 4.)
Validation (determinations of accuracy) of the two approaches’ facility staffing estimates in terms of the safe and efficient performance of controllers at facilities that meet their staffing targets would help determine the preferred approach. However, the accuracy of the two modeling approaches
could not be determined with data available to the committee. Data available to the committee that could have been used to help validate the two modeling approaches, such as overreliance on overtime at understaffed facilities, are not a helpful indicator of workforce strain for this purpose because of the extent of its use in facilities that are not understaffed. Safety indicators not available to the committee, such as the frequency of confidential data on close calls, including voluntary employee reports on fatigue, and other safety metrics might be useful in validating the staffing adequacy of the 40% of facilities that are within +/−10% of their staffing targets as estimated by the traditional approach. This could be done by comparing the frequency of close calls and other safety metrics at the facilities within +/−10% of their staffing target with those that are more than 10% below or above their staffing targets. (See Chapter 3.) Although the committee could not validate either approach, it believes the traditional facility staffing modeling approach has more advantages and fewer weaknesses than the collaborative approach.
It is difficult for even finely tuned facility staffing models to account for all unique operational conditions at individual facilities, such as the mix of traffic, specific training requirements, and facility layout. FAA addressed this problem in the past by setting facility staffing targets that combined modeled estimates with input from the field regarding unique operating conditions. However, past processes of gathering field input lacked transparency and consistency; these could be improved using a survey approach such as the one developed by the collaborative method. (See Chapter 4.)
Recommendation S-4: FAA should return to the practice of adjusting facility staffing targets estimated with its traditional approach with input from the field for facilities with unique operating conditions by using a structured survey process such as the one used in the collaborative modeling approach. This process should be transparent, consistent, and reflect site-specific factors. (R4-3)
Recommendation S-5: FAA should reexamine its task load requirements models and whether its estimated facility staffing standards provide adequate time for other duties, arrange for any necessary updates, and establish a system to periodically evaluate the need for future updates. (R4-1, R4-2, R4-4, and R7-1)
Recommendation S-6: FAA should use the confidential and other safety indicator data that it collects to validate, to the extent possible, its traditional facility staffing models. (R3-2)
ASSESSMENT OF FAA’S RESPONSES TO FATIGUE AND SAFETY RECOMMENDATIONS OF NASEM 2014 REPORT
The NASEM 2014 report made 26 recommendations, FAA’s adequate responses to most of which are discussed throughout the report. The most important ones not fully addressed regard the fatigue associated with understaffed facilities that could compromise safety. Fatigue among controllers has also long been compromised by most facilities’ use of weekly fatigue-inducing shift schedules that rotate backward against circadian rhythms. Many controllers prefer these schedules because they compress their workweek and give them at least 80 hours off between weekly shifts. The NASEM 2014 report cautioned against using such shift patterns because they do not allow for sufficient restorative sleep, particularly before the beginning of the midnight shift. In response to the recommendations of an FAA expert panel’s 2024 recommendations, which parallel those of the NASEM 2014 report, FAA mandated longer recovery periods between shifts in FY 2024, which will occur in FY 2025, and initiated the elimination of backward rotating shifts, which is expected to occur in FY 2026.
Due to lack of the research recommended in the 2014 report regarding the relationships between safety risks and facility staffing levels, the committee for this report is unable to isolate the effects of past widespread use of counterclockwise shift schedules on fatigue from the effects of understaffing. The 2014 NASEM report committee made two other recommendations that were also made by FAA’s expert panel on fatigue to (a) manage fatigue levels inherent in 24/7/365 operations through a robust Fatigue Risk Management System (FRMS) and (b) facilitate shift schedules in conformance with fatigue rules and principles using efficient shift scheduling tools. The issue of inefficient shift scheduling emerged late in the committee’s study, which did not allow time for a thorough analysis. The analysis the committee was able to conduct indicates it is not an isolated problem. Inefficient scheduling may be related to perplexing trends in regular hours worked, TOP, and increased overtime mentioned above. The inability of FAA to implement a robust shift scheduling software package it acquired and began testing in collaboration with NATCA in FY 2012 means that facilities currently lack a robust shift scheduling tool capable of enforcing fatigue rules while also making the most efficient use of available staff. (See Chapter 3.)
Recommendation S-7: FAA should fully implement previously recommended research on the relationship between staffing levels and safety risks, adjust its facility staffing model parameters accordingly, and implement the recommended robust FRMS and efficient shift scheduling tools. The latter should include rigorous requirements and Request
for Proposals process that includes a build versus buy decision for shift-scheduling software. This process should also include implementation and organizational change management support for greatest likelihood of successful implementation and adoption. Congress should provide any additional funding required to increase FAA expertise in FRMS and for it to acquire (or develop) and implement the recommended shift scheduling tools. (R3-1, R3-5, R3-6, and R3-7)
This page intentionally left blank.
