There have been several things written over the last couple of years about working conditions in Amazon fulfillment centers. (See, for example, here, here, and here.) Now we have a BBC report complete with hidden-camera video of what it is like inside a fulfillment center.
If you prefer to read, you can also check out “Amazon workers face ‘increased risk of mental illness’” (Nov 25).
Working in an environment that combines tedium with constant monitoring is going to be stressful. That would be true whether it is a fulfillment center, a call center, or on an assembly line. It may be more surprising that a fulfillment center is so stressful since the job description amounts to “walk around and pick up stuff” as opposed to “install the same thing over and over again on 400 cars per day.”
Of course, that doesn’t justify picking rates that end up wearing people out. But it is not clear what constitutes a reasonable rate. First, I would wager that if the target rate was set at 80 per hour (which is in the range that undercover reporter was achieving), the average worker would pick more per hour than if the rate were set at 70 per hour. That is, workers are going to respond to the target in front of them. Given that, how should they the target? Suppose Amazon has a lot of data and knows that distribution of items picked per hour ranges from 60 to 120. Should it pick, say, the median and expect that pickers will only hit it half the time? Or should it aim higher and know that pickers will hit the target only, say, 25% of the time? That question doesn’t have an obvious answer even if worker performance is independent of the target. Once effort goes up with the target, aiming for a higher target even if the average worker only hits it a small fraction of the time becomes an attractive choice.
Further, Amazon may not care that much about the average worker. They have sharp seasonal swings in their business so relying on a mix of permanent employees and seasonal employees is a must. If workers differ in their ability to pick orders, Amazon should favor moving proven, high-productivity workers to permanent status. High targets are then a way to learn who can consistently work at high rates.
Still, there is no question that there is something unseemly about these reports of Amazon working conditions. It is not a job I would enjoy. It may just be the cognitive dissonance of pairing high tech with physical exertion or my convenience with someone else’s sweat.