In the basement supply room of Children’s Hospital of Philadelphia, where the diapers, IV bags, and baby formula are kept, the numbers call the shots.

There are people who enforce the numbers, yes, but it’s the numbers that reign supreme.

Two hundred and fifty lines per shift, 1.5 hours per pallet. And the most mysterious: 1.2.

That 1.2 is the productivity rate that the warehouse pickers — workers employed by CHOP subcontractor Canon Business Process Services who unpack boxes of supplies, put them away, and pick them off shelves to fill orders that come in throughout the shift — are supposed to meet as a group. Last year, the rate, seemingly related to worker speed, was a little easier: 1.5. In 2017, anything over 2.0 was “extremely bad,” said a former supervisor.

It’s not clear to them what their rate is while they’re picking or how it’s calculated — the former supervisor, who spoke on condition of anonymity because he works in the same job at a different hospital, said it was “computer-generated.” What workers say they do know is that the rate is nearly impossible to meet if any little detail is off: If they’re understaffed or overstaffed, if it’s a holiday, if there’s a person who’s new and just getting up to speed.

And they know that at the end of the shift, if the computer says they picked more slowly than expected, they’ll get written up.

The result, four current and former staffers said, is a workplace where workers live in fear of getting fired because of a number that doesn’t account for the realities of the job, a number they barely even understand.

“It’s exhausting if I’m killing myself trying to meet these invisible numbers,” said a picker who works the overnight shift and spoke on the condition of anonymity for fear of losing his job.

Welcome to the Amazonification of work. In recent years, “rate” has become the new gold standard for running a warehouse, made prominent by Amazon and numerous accounts of its workers.

“Rate was constant,” one worker in California told Gawker in 2016. “Rate was on you all day.”

Companies have developed technologies to track productivity at a scale and granularity beyond their previous dreams, and they use this data to increase efficiency, make their business more attractive to investors, and control their workers.

Now, quotas go far beyond the plight of the desperate salesmen whose jobs are in jeopardy in David Mamet’s Glengarry Glen Ross:

• At the Philadelphia Marriott Downtown, housekeepers can’t go home until they clean a predetermined number of rooms, even though, housekeepers say, rooms vary in cleanliness depending on such factors as length of stay and whether guests declined service during their stay.

• At some retail chains, workers say they have to persuade a certain number of customers to share their email addresses or open store-brand credit cards, or else face a cut in their work hours.

• UPS uses sensors to track its drivers’ stop times, backup speeds, and seat belt use. The company is not allowed to fire workers based on these numbers, but only because the Teamsters, the union that represents UPS drivers, fought for that language in their contract.

Limits to workplace tracking are ‘blurry’

Ideally, collecting data on how a business is run and using it to develop efficiencies is a win-win for a company and its workers. Market forces in such industries as health care make data-driven analysis necessary, experts say, and certain business functions — such as a hospital’s storeroom — will be more prone to it than others.

“There’s always going be pressure to reduce costs and be as efficient as possible in any department that’s not providing direct patient care,” said Andy Murphy, a Chicago-based health-care consultant with FTI Consulting.

And although it’s largely accepted that employers have a right to track the performance of their workers, the problem, said Alexandra Mateescu of the research institute Data & Society in New York City, is that the boundaries of that kind of tracking are “incredibly blurry, both normatively and legally.”

“It’s very hard to point to an authority that says, ‘Here are the proper limits of tracking in the workplace.' Often, these boundaries are set within the industries themselves," said Mateescu, whose research focuses on labor and technology.

Canon spokesperson Ken Neal said the company would look into the situation at CHOP but declined to comment further. The company “deeply respects and values our employees,” he said in a statement. CHOP said it had no comment.

Quotas as a replacement for training

One former Canon worker at CHOP said that when he was hired last summer to be in charge of the carousel, which stores such items as toothbrushes and booties, his supervisor told him managers would be patient as he learned what they said was the busiest and most important position on the floor: the “quarterback of the night,” he said his supervisor called it.

Which is why he was surprised when he got written up in his first month for failing to meet productivity goals.

The worker, a 23-year-old who lives in West Philly and now works at a school, said he was hustling to learn about the hundreds of items on the carousel, all the different types of diapers and syringes. (Each type of item, regardless of how many are ordered, constitutes one “line.”) He got two 15-minute breaks during his overnight shift, but he said he realized he couldn’t afford to take them. He knew he could move faster unpacking boxes of supplies and setting them up on the carousel’s shelves, to which he’d later return as orders came through, but that would mean sacrificing accuracy for speed — and he didn’t want to be sloppy.

He continued to get reprimanded, which involved supervisors pulling him off the line to discuss the write-up. And every time he was brought in, he lost more time toward making rate. “Do they even think about that?” he wondered.

Six months after he was hired, he got fired for what supervisors described as unexcused absences.

His experience speaks to a pitfall of quotas described by Mateescu: Are they being used as a solution to the more time-intensive work of vetting and training workers?

“There needs to be a way to account for a learning curve,” she added.

Why quotas are common in subcontracting

In documents relating to an open 2018 Philadelphia Commission on Human Relations complaint brought against Canon by employee Edward Osborne, Canon director of employee relations Edward Sample made references to Osborne’s lack of compliance with “CHOP’s performance metrics," not Canon’s.

Osborne’s supervisors told him that he “must improve his performance metrics and meet CHOP’s expectations," Sample’s statement read.

This kind of standard-setting and monitoring is common in subcontractor relationships, labor expert David Weil wrote in his 2014 book, The Fissured Workplace, because a company otherwise risks tainting the reputation it’s built when a contractor does not deliver. Performance metrics, in other words, are a way for companies to control their contractors, just as the contractor uses them to control its workers.

Fred Landgraf, a Florida-based supply chain expert with health-care-focused Intalere, said hospitals have been slow to modernize their supply chains.

They are just beginning to hold people accountable to productivity rates “because of the pressures that health care is facing across the board," he said. Fifteen years ago, no one knew how many lines their workers were picking, but today the standards are getting tougher, he said: 20 lines per hour used to be the norm, now it’s 40.

These changes are largely positive, he said, as hospital supply chains become better managed and more effective. But he said it’s important to remember that “you’re still dealing with people.”

There are constraints that workers don’t control — for example, if there’s more demand at certain times on certain days. In those cases, managers should recognize the patterns and adjust staffing to meet demand.

But, he said, “you have to do that humanely, for lack of a better term."