Last month, the labor union representing American air traffic controllers called for congressional hearings into staffing shortages at the Federal Aviation Administration (FAA). According to the National Air Traffic Controller’s Association, the FAA has failed to meet its hiring goals for the last five years, which has lead to personnel shortages across Georgia, Illinois and New York. In Texas, staffing at the regional air traffic facility is nearly 45 percent lower than believed necessary. Given that the nation’s controllers are responsible for the lives of nearly two million passengers a day, action certainly seems warranted.

However, controllers come with a hefty price tag. Their salaries, which average some $136,000 annually, are among the highest in the federal government, chewing up some 18 percent of the FAA’s budget. All in all, controllers cost the taxpayer nearly $2.8 billion annually. It is therefore important that staffing decisions be driven by sound analysis. And this means asking the right questions.

For example, how many controllers are required to staff the nation’s skies? While some claim controller staffing is down, an analysis by the Inspector General’s office found the size of the workforce to be “relatively constant” between 2000 and 2012; this despite the fact that air traffic had dropped by 23 percent. A more recent study by the respected National Academy of Sciences found some of the forecasts used to guide controller hiring to appear “unreliable” and that “persistent optimism (in those forecasts) brings about a tendency to hire more controllers than required.” Academy researchers concluded that the process of calculating staffing ranges was “opaque” and “open to concerns that they were arbitrary or inconsistent.” This raises questions as to whether or not controller ranks are overstaffed to begin with.

Equally important is the question of workforce productivity. Although critics of current staffing levels contend that safety is not at risk, that safety is purported to come at the cost of efficiency and modernization. This means less than ideal productivity. Increasing staffing should provide relief by fostering scheduling practices more favorable to maximizing productivity. However, an audit by the Transportation Department found that the FAA does not have ways of measuring the effects of these practices. Investigators specifically noted the “FAA’s inability to reach consensus on which metrics should be used to measure controller productivity.” Given such findings, how can the envisioned benefits of more controllers be shown to the American taxpayer?

Last but not least is the issue of controller fatigue. Current scheduling practices allow controllers to work five 8-hour shifts over 88 consecutive hours. In comparison, the average person will perform the same amount of work over 104 hours. The practice is popular because it gives controllers three days off before they begin their next shift. It is also concerning because it has historically provided controllers with a mere eight hours of relief between a day shift that ends at 2pm and a night shift that starts at 10pm. Researchers from the National Aeronautics and Space Administration have found that controllers begin their night shift with just 3.1 hours of sleep in hand and are more likely to catch themselves “about to doze off” while working. Given that the health of the skies is tied to the health of the controller workforce, would more staff be used to eliminate current scheduling practices?

Raising such questions should not discount the legitimacy of requests for more staff. With over 30 percent of the workforce already eligible to retire, more controllers will almost certainly be needed to fill the spots created by retirees. The question is how many more.