Most of the time, it’s a cop-out to blame technology for the human misbehavior that it enables. It isn’t PowerPoint’s fault that your co-workers add too many slides to their presentations. It isn’t Facebook’s fault that “friends” whom you barely know make odd comments on your photos. It isn’t Auto-Tune’s fault that Paris Hilton thinks she’s a singer.
But when the stakes are much higher, even software engineers should do some soul-searching. Late last month, just as shoppers around the country were girding for Black Friday, the San Francisco Board of Supervisors approved a “retail workers bill of rights” designed to give workers at retail chains more predictable schedules and discourage last-minute scheduling changes. It was a direct response to a powerful new employment trend: Increasingly, major retail and restaurant chains fine-tune their staffing — and hold down labor costs — via sophisticated software that looks at a store’s past performance, weather patterns, and real-time sales data.
The software plays an integral role in so-called just-in-time scheduling systems, which help ensure that a store won’t have eight cashiers working when there’s only enough business for four. For workers, though, these systems have serious downsides: irregular shifts, significant schedule changes on short notice, and huge variations in hours from week to week.
Earlier this year, The New York Times profiled part-time Starbucks barista Jannette Navarro, a San Diego single mom who couldn’t arrange child care or take classes because her hours fluctuated so wildly. Workers at chains from Walmart to Jamba Juice have gone public with their frustrations. “These hours don’t match the basic realities of people’s lives,” said Carrie Gleason, director of the Fair Workweek Initiative at the Center for Popular Democracy. The burden for workers with families is particularly heavy, she added. “Kids need routine, but when you work in retail routine doesn’t happen.”
The market leader in this workforce-management industry is Chelmsford-based Kronos Inc.; other players include SAP, ADP, and Oracle. What these firms have to decide is whether their products can be a force for greater equity in the workplace — or will remain one more way, in an uncertain economy, to shift more of the risk onto low-wage employees with little leverage.
Strikingly, none of the researchers or labor advocates whom I contacted blamed Kronos or its competitors for schedules that, ultimately, reflect the employer’s values. Still, all the evidence suggests that relying on faceless algorithms makes it easier for employers to casually jerk workers around.
If you worked in retail 20 years ago, your manager would post a handwritten schedule on the back of the bathroom door every week or two. She might have expected you to work every other Friday night, because spreading unpopular weekend shifts around helps morale. If you worked Tuesday and Thursday last week, she might give you the same shifts this week, because reinventing the schedule from scratch would be a hassle. She made judgments about which inconveniences you might grumblingly accept — and which ones were too burdensome to demand.
A robo-scheduler doesn’t recognize such objections unless it’s programmed to. “The algorithm did it,” a manager might rationalize — especially when headquarters is keeping a close eye on staffing at every store.
It also can’t be a coincidence that, as scheduling software proliferated over the last decade, so did the widespread use of on-call shifts, which require part-time employees to check in an hour or two beforehand to see if they’re needed. For workers who need to arrange child care, or work a second part-time job, being called in with a couple of hours’ notice is a disaster.
The workforce-management industry is a little cagey about how its products affect workers. “Reduce labor costs by efficiently scheduling, monitoring, and managing your workforce,” promises an Oracle marketing website. “Provide outstanding customer service as you control labor costs,” says the pitch for a Kronos product . But in a recent interview, Kronos’s vice president for business development, Charles DeWitt, argued that minimizing labor costs is “an afterthought in the calculation.” As he tells it, it’s hard enough just to match the mix of skills and certifications that a retailer needs at a given time (fluency in Spanish, the capacity to perform certain management duties) with the availability of workers, whose time constraints vary greatly.
In our conversation, DeWitt sounded genuinely interested in addressing some of the problems worker advocates have raised. Kronos is developing metrics for how often the hours an employee works differ from what’s on the original schedule, how well staffing respects workers’ preferences, and how widely a given employee’s hours vary from week to week. This is encouraging, but also unsettling: Shouldn’t such considerations have been part of the equation all along?
Eventually, labor laws have to adapt as well. Earlier this year, US Representative George Miller of California introduced the Schedules that Work Act, which would compensate retail, food service, and janitorial workers for last-minute schedule changes. But nobody thinks the federal legislation will pass anytime soon. If history is any guide, liberal states like California and Massachusetts will enact some controls, while other states will blow the issue off entirely.
Worker advocates need to look for additional pressure points — and software makers ought to own up to their own role in creating the current system. If technology firms and the retailers who hire them are looking at the right data, over a period of time that extends beyond the current quarter, they’ll be able to verify what labor activists have long believed: that more stability for workers reduces turnover and improves customer loyalty.
Maybe it’s too simplistic to hope for a simple software tool that allows employers to upgrade their schedules from 1 (“sadistic”) to 10 (“workers’ paradise”). If nothing else, Kronos and its competitors can help simply by confronting retail chains up front with the sacrifices they’re expecting from their workers.
Related:firstname.lastname@example.org. Follow him on Twitter @danteramos.