Amazon is a gigantic and rapidly growing company. It’s expanding its labor market footprint in all kinds of ways, and one of those ways is its enormous warehouse operation that offers job opportunities that are, by absolutely all accounts, incredibly arduous.

You could do a lot of different takes about the prospect of a bunch of really difficult blue-collar jobs arriving in town, but I read a new one this week in The Washington Post in which Abha Bhattarai quoted an analyst from the National Employment Law Center who maintained that the arrival of an Amazon warehouse in town actually reduces wages.

“Once Amazon gets to town, turnover skyrockets and wages decline,” Bhattarai quotes Irene Tung of NELC saying. “Amazon is the standard-bearer, and it is dragging down working conditions for everybody.”

Is that possibly true? Is Amazon itself not just a rough place to work, but its existence actually reduces pay and working conditions for people who don’t work there?

It’s an important question because the phenomenon of some jobs being more dangerous, more back-breaking, or otherwise more demanding than others is not new. Work in the logging and fishing industries, for example, is much riskier to the workforce than normal food service or retail jobs. But logging and fishing pay unusually well because nobody would take a job like that unless there was a pay premium. The technical term for this is “compensating differential” — the extra money an employer needs to offer to persuade people to take jobs that have unusually bad non-wage characteristics.

Whatever else you may say about compensating differential industries, their existence is generally unambiguously good for the people who don’t work in them. A fast-food restaurant in an area with a substantial mining industry is going to pay better than a similar restaurant in an area without one because it needs to compete with the mining operation for workers. Some people take the more arduous job in exchange for the higher pay, but even for those who ultimately don’t, the existence of the option is a net positive.

Tung’s charge is that Amazon is somehow different and its warehouses not only offer tough working conditions but somehow make working conditions elsewhere worse. If that’s true, it obviously has important policy implications. But the evidence seems to indicate to me that this is not true at all.

What’s true — Amazon warehouse jobs don’t seem great

In the current internet climate, it seems like all of the commentary is very hostile to Amazon, and if you ever say anything good about it, you are inevitably going to get yelled at.

So it’s worth a quick reality check. Last week, the Verge published its annual survey on public perceptions of the technology industry, and views of Amazon are almost uniformly positive.

What’s more, confidence in Amazon is growing over time. In a separate question, 27% of respondents said they’d grown more trusting of Amazon in the past year versus 14% who are less trusting.

So my generally favorable attitude toward Amazon (and yours, if you are honest with yourself) is extremely normal, and the showy Amazon hate that you see everywhere online is weird and idiosyncratic.

That being said, it seems like working in an Amazon warehouse is really unpleasant. That’s what Carole Cadwalladr found in 2013, and it’s what Ryan Fan found in 2020. People I’ve spoken to who’ve worked at Amazon in high-level white-collar positions describe it as a grueling, wouldn’t-want-to-do-it-again experience, so it’s no surprise that the blue-collar work is even more mentally and physically taxing. This snippet from a Mental Floss piece sounds kind of dystopian:

Working at an Amazon warehouse can be physically challenging, with lots of bending, lifting, and moving. As a result, their vending machines offer more than just candy bars and potato chips. “We have no-cost medical vending machines,” Alex, a packer at an Amazon warehouse, tells Mental Floss. “They hold individual dosage packets of things like Advil, Tylenol, Tums. I've used them before when I ran out of my own stash that I bring in my bag, and they come in handy.”

One of many, many New York Times exposés of working conditions at Amazon underscores that over and above the arduous nature of the task, the management of the company’s vast blue-collar workforce is harsh:

In contrast to its precise, sophisticated processing of packages, Amazon’s model for managing people — heavily reliant on metrics, apps and chatbots — was uneven and strained even before the coronavirus arrived, with employees often having to act as their own caseworkers, interviews and records show. Amid the pandemic, Amazon’s system burned through workers, resulted in inadvertent firings and stalled benefits, and impeded communication, casting a shadow over a business success story for the ages.

And this is not just a question of media takes. Amazon has incredibly high turnover in these warehouse jobs. They offer decent money, so a lot of people want to apply, but the working conditions are pretty bleak and people tend to move on quickly. Amazon’s official line is that a lot of their churn is due to people coming back sporadically to work for the company when they want to earn more cash.

Is that actually true? It’s surely true for some workers. After all, in addition to the baseline large size of the Amazon workforce, there’s a predictable seasonal spike around the holidays. For all kinds of jobs with that pattern, there are always some “regulars” who for one reason or another like that irregular work schedule. We all saw Nomadland. At the same time, Amazon hasn’t provided the world with hard numbers on this, which I think they would if the data were overwhelmingly persuasive.

There’s a theory — outlined by David Leonhardt back in June — that the company philosophically favors a high-turnover workforce; that in their mind it’s like the Army, where the vast majority of the soldiers do a tour or two and then leave, by design. A company that feels ambivalent about the basic concept of employee retention is inevitably going to be an unpleasant place to work. But at the same time, it’s not like Jeff Bezos found a loophole in the rules of the universe and doesn’t need to care about job quality at all.

Amazon responds to basic incentives and pays well

You’ve probably heard that rather than spend the money needed to install air conditioning at its warehouses, Amazon responds to employee overheating by stationing an ambulance outside.

And when Spencer Soper reported it 10 years ago for the Allentown Morning Call, this was true. What’s also true is that 10 years ago, the overall labor market situation in the United States was really bad. How bad? Well, even though the Great Recession had been officially over for more than a year, the unemployment rate was still higher than its average during the plague year of 2020.

In other words, it was a great time for a company looking to expand its blue-collar workforce without treating its workers very well. But as Amazon kept growing and the economy improved, the company responded to the bad press by adding air conditioners.

By 2018, Amazon had boosted its entry-level wage to $15/hour. This spring, they lifted it again to $18/hour. According to the Bureau of Labor Statistics, the median hourly wage in the United States last year was $20.17, so $18/hour is very good for a job that doesn’t require any experience or qualifications.

None of this is to say that Amazon is run by a bunch of sweethearts. But they need to get people in the door, so they offer good pay and try to address the most egregious complaints raised about them in the press. It’s hard to see how they could be lowering wages by entering markets.

The warehouse sector is growing nicely

I haven’t been able to find a study precisely on the spillover impact of Amazon opening up a fulfillment center in a town. But Ellora Derenoncourt, Clemens Noelke, and David Weil did a study that looked at local spillovers from Amazon wage increases, and they found (unsurprisingly) that when Amazon raised pay to $15/hour, that raised wages elsewhere in communities that have Amazon warehouses.

But could it be that Amazon somehow shows up and (with its cutthroat practices) drives other warehouses out of business, leaving local communities worse off in the aggregate? In theory yes, but what we see in practice is that the warehousing industry has been growing at a nice clip.

And pay in this field, in particular, has been rising a lot recently. These days it’s important to adjust for inflation, which has been high, but you can see that in transportation and warehousing, nominal pay has been growing faster than inflation, so it’s hard to see what downward pressure you could possibly attribute to Amazon.

Indeed, one line of criticism mounted against Amazon is that their relatively high pay is just another shady monopolist tactic:

“Being a business major, by raising workers’ pay during the virus lockdown, it has economic implications,” the same former Inland Empire Warehouse worker and SDSU student said. “It might make it tougher for other warehouses and smaller companies to hire people. When Amazon raised their salaries, it made them more of a monopoly.”

I think a realistic view of the situation is that it’s just the reverse. There is a lot of competition in the retail industry, and Amazon engages in that competition vigorously. They have a pretty cutthroat attitude toward their workforce, but they are also trying hard to expand and respond to real-world conditions. All signs point to their growth pulling things up, not dragging them down.

Isn’t this what we were nostalgic for?

It seems to me that fairly recently, the whole country was tied up in arguments grounded in nostalgia for the old industrial economy and its factory-floor jobs.

Skeptics could, of course, note that those jobs were a lot more dangerous and involved more physical exertion than working behind a cash register at a shopping mall. But in exchange for those downsides, factory work had some appealing features including relatively predictable shifts and relatively high pay for people with less education. I also think the fact that they didn’t involve interpersonal service was appealing to a lot of people, especially men. There’s no emotional labor on the factory floor — you just need to show up and do your job.

The new warehouse economy has basically all of those attributes — both good and bad — while being almost purely additive to the existing in-person service economy.

Conor Sen wrote a piece about this that upset people due to a headline reference to “factory towns.”

People like to get mad about Amazon on the internet, and in this case, they seem to have been confusing the idea with the concept of a “company town” where your boss is also your shopkeeper and your landlord.

But he’s just saying that we used to have lots of smaller cities that were anchored by a major industrial employer. When those employers went away, there were a lot of downsides. The future, however, might be a major logistics employer that serves a similar function. This really does seem to be a problem for some small businesses who are struggling to compete with Amazon on the labor front. But there are way more people who participate in the working class labor market than who own small businesses. This kind of trend is a positive for the country and the world.

Amazon deserves scrutiny because, as we saw with the air conditioning, concrete reporting on specific issues can lead to changes. Amazon is so big that improving pay and working conditions is a powerful lever for broader change. And blue-collar Amazon work has a lot of downsides. But so do virtually all blue-collar career paths in the contemporary United States, and the idea that Amazon’s expansion is making things worse has basically no evidence in its favor, and it seems contradicted by both common sense and the available research.