Air Traffic Controllers: the Canary in the Coal Mine?
Air traffic controllers sleeping in their towers are emblematic of a much larger problem: how employers hurt American competitiveness through scheduling practices that create a bleed of back-end costs.
“They fall asleep because of the shifts they work,” said a controller in a letter to the San Francisco Chronicle. A typical shift pattern starts at 6 a.m. on Monday, 10 a.m. on Tuesday, 1:30 p.m. on Wednesday, 8 a.m. and then again at 10 p.m. on Thursday. “That’s right, two shifts in 24 hours. Which means that at 5 a.m., most controllers are exhausted,” concluded the letter writer.
The air traffic control issue reflects a larger problem in the American economy, as documented in studies by Susan Lampert, Julia Henly and Anna Haley-Lock. Many private employers have shifted from the stable schedules of yesteryear to “just-in-time” schedules that change, with very little notice, from day to day and week to week. Airline catering workers are sent home if a flight is canceled, but are expected to work overtime on busy weeks. Nurses’ aides arrive to find that their shifts have been canceled because the patient census is lower than expected. Restaurant staff often are sent home if the ratio of labor costs to sales staff exceeds 29% by 3 p.m., or if that ratio seems unlikely to drop below 21% by the end of the business day, according to one study.
Sounds scientific. It’s not. Just-in-time schedules take into account neither basic human realities, like the need to sleep, nor demographic realities. Such schedules don’t fit with Americans’ family caregiving responsibilities. Workers in other industrialized countries have paid family leave and mandatory vacations, as well, often, like limits on overtime and subsidized child care — family supports not mandated here. As a result, one in four American families handles child care by tag-teaming (where mom works one shift, dad works a different shift, and each cares for the children while the other is at work).
This makes mandatory overtime a problem. In the U.S., grandparents often tag team with their children to care for grandchildren: about half the managers in one study reported having at least one employee caring for family members other than their own children. In some inner cities, grandparents are the primary guardians of 30% to 50% of children under 18, according to a study in The Gerontologist by R. Pruchno.
The relative’s hourly workers rely on often have just-in-time schedules themselves.
This leaves employers competing with their employees’ relatives’ employers. In addition, when just-in-time schedules give only part-time hours to a large pool of workers (which is common), employers are left competing for workers’ attention with their employees’ other employers.
Americans’ heavy reliance on tag teaming and other forms of family care combines with just-in-time scheduling to drive up labor costs through sky-high turnover and absenteeism. In addition, key aspects of just-in-time scheduling are illegal in other countries. In Canada, for example, employers are required to pay for at least eight hours of work if an employee reports for a scheduled shift.
Canceled shifts and erratic schedules produce a pattern of serial quitting that costs American employers dearly. Annual attrition over 80% is commonplace in just-in-time jobs, with attrition rates as high as 500%. Given that replacing a single hourly worker costs between 30% to 75% of the annual salary, this is just bad management. It’s bad management, too, when 80% of employees are on probation due to tardiness and absenteeism, as was the case in one flagship department store studied, according to a 2006 study by Julia Henly, J.R. Shaefer, and Elaine Waxman.
The solution is schedule effectiveness, a concept pioneered by workforce consultant Lisa Disselkamp. Scheduling needs to come out of the 19th century — a paper and pencil world where the assumption was that any employee worth having was always available for work. Ninety-four percent of store managers try to hire for “open availability” — the ability to work anytime — according to one study. It’s unrealistic to design 21st-century schedules around the assumption that anyone worth hiring has someone else taking care of their children. Sorely needed is a more scientific scheduling process that uses “cloud” technology, which now offers online scheduling for as little as $1.25 per employee per month.
Scheduling effectiveness starts with a survey to identify common scheduling constraints and to build them into the basic schedule.
The next step is to analyze past schedules to find hidden scheduling stability: one study found that 80% of the hours worked in retail stores were stable, month after month. (Employers were unaware of this.) The third step is to lengthen the period within which supervisors can “stay within hours,” so that if labor demand is lower than expected, supervisors can achieve the required ratio by the end of the week by not replacing someone who calls in sick rather than by sending home people who have already reported for work. The final step is to determine the optimum advance notice of employee schedules. This needs to be done empirically: by testing out various notice periods and comparing the costs associated with more notice with the savings achieved by lengthening the notice period.
Effective schedules are key to American competitiveness.
The best way to solve the airport slumber-party problem may not be to schedule two controllers on overnight shifts, which is one of the proposed solutions. A cheaper alternative is to redesign schedules so controllers can get enough sleep. The larger goal is to identify scheduling equilibrium: the point at which one cannot drive front-end labor costs (a tight fit between labor supply and labor demand) any lower without driving up back-end labor costs (turnover, absenteeism, etc.). Effective scheduling can keep us safer in the air and in hospitals, allow for better service in retail stores, hotels, and restaurants, and enhance the overall competitiveness of American business.