How Do You Bargain With an Algorithm? The Hazards of Living in Uberland
11.08.2018

Alex Rosenblat’s new book Uberland: How Algorithms are Rewriting the Rules of Work chronicles the utopian rhetoric and complex reality of the gig economy, particularly its computerized management of drivers. We talked about the problems of today’s platform economiesand whether blockchain can change the math. Rosenblat is a researcher with Data and Society, a nonprofit institute focused on technology issues. This interview has been condensed and edited for clarity.

Just in time for publication of your book, Uber has been valued at $120 billion. What does that say to you as somebody taking a hard look at the company?
It says to me that someone’s going to earn a lot of fees [laughing]. But no, it’s hard for me to comment on their valuations, because they change over time. I don’t think we’ll have a clear view of what the company is worth until their IPO.

I guess what I meant was…that’s a lot for Uber.
Yeah, that’s a lot of billions. Uber’s outsized valuation is partly how it’s become such a dominant symbol in the future of work. It has become this steady media beat where labor has become a tech beat for journalists. It’s blurred boundaries in ways that make it representative of so much.

Uber seemed like a less predatory form of commerce after the subprime mortgage crisis left cities covered in blight.

It’s like how bitcoin has turned finance into a tech beat. Bitcoin and sharing economy both grew out of the Great Recession, both rhetorically and chronologically. Did a loss of faith in institutions also play a part in Uber’s story?
I think that the recession made everybody very receptive to what was termed the sharing economy. When I started studying Uber in 2014, it was still a darling of Silicon Valley. It was celebrated because it was offering something really valuable. Uber said to a country that had just faced foreclosure crises and people losing their jobs, “Hey, we’ve got this technology, and it connects people, and we can actually get people to share their underutilized assets and skills.” Which seemed like a less predatory form of commerce after the subprime mortgage crisis left cities covered in blight. It was more appealing.

And Uber also said, “Look, we’ve got something that draws on a cultural legacy in America, this dream of entrepreneurship—to be your own boss, create your own business, to become very successful in that way. And we can scale entrepreneurship for the masses.” So I think it gained power in a moment where it was able to use that legacy of cooperative commerce, and a desire to recenter the economy around a more moral framework.

Like bitcoin, Uber also grew by either breaking the law or using technology in ways that worked around the law. Should we be rethinking government so it deals better with this arbitrage of the rules?
I learned a really interesting phrase recently: regulatory forbearance. It means when the laws exist, but they’re not enforced. What Uber and bitcoin have done is put a pause on some of the regulatory barriers because they offered something that’s quite popular, and kind of interesting, and might be innovative.

There’s this moment after Uber arrives in a city, where it’s operating in a gray zone. It’s recruiting drivers, and providing a popular service to consumers who suddenly have a more efficient way of getting around. It’s got a foothold, and regulators can try and crack down on Uber. But Uber can use the politics of consumer choice to reduce legal challenges.

It makes a compelling argument that makes it difficult for governments to enforce existing regulations. And it’s created an opportunity for older industries to take note and say, ‘ok, should we adopt these innovations?’ Do we need to be regulated in the first place? Do we need all these different regulatory regimes to evaluate drivers, or is a rating system enough? And it seems self-evident that this is what the people want, and are willing to do the rating.

It sounds like you’re sympathetic to the argument that either less regulation, or breaks in regulation, can spur innovation that ultimately does lead to a better world.
Here’s where it gets tricky. Let’s say you do have a successful, innovative new service and it disrupts an old industry and consumers are happier. What happens when that service starts to have the power of a monopoly? That’s a very different question than, ‘can I get a ride quickly?’

And that’s the question that’s going to be looming. Some cities, for instance, think this very appealing [ride-hailing] technology means they should invest less in public transit, and maybe subsidize ride hail services instead.

But what happens when there’s no public transit and Uber or Lyft decide to pull out one day? That happened—they left Austin, Texas. Uber and Lyft had recruited drivers who started to rely on this income for their livelihoods. And when the city tried to pass some regulations, such as requiring drivers to undergo fingerprint-based background check, they left town.

In the book, you unpack the term “sharing,” and how that developed. Do you see any parallels in how blockchain technology is discussed? Are they both examples of terminology that hide a more complicated truth?

What they [both] did was inspire the public imagination. Regardless of what kinds of trouble they are entangled in on the ground, they created something really cool. They were just like, ‘hey, why don’t we think about this another way?’

There was an idea that blockchain could decentralize power and create transparency in the sharing of information. And that’s pretty interesting. Maybe this is a way of achieving transparency in government, or maybe we can build a new surveillance system where all transactions are recorded, and there’s no one central authority that can enact unequal power. I think it draws on similar ideas to how the sharing economy was promoted.

But I wonder if everything should be recorded in a public ledger. Should someone’s visit to an abortion clinic be logged on a ledger? These technologies are invested with a different sense of how we should govern society, and that gives me pause, because they might disadvantage marginalized populations, or they might expose privacy in ways that could have pretty bad effects.

How do you view the moral arc of the idea of the sharing economy? Did it start with genuine utopian idealism before becoming this more complicated reality?
The sharing economy rhetoric leveraged our sense of communal, reciprocal forms of transactions, and used it to mask some of what was really going on. There have been some efforts to salvage some of the promise of the sharing economy, such as efforts to create platform cooperatives—services like Uber, but that are owned or controlled by the workers and other stakeholders.

Platform cooperatives—that’s a mouthful. Are there successes there? Or is it all kind of in development at this point?
My takeaway is that they’re not as scalable and adaptable as the platforms we have. Part of Uber’s value is that it can operate at scale. You can disembark in some city in the middle of the country and use the same app to get a taxi. The global user-base is part of what makes it unique, and more adaptable than a co-op, which tend to operate better on a small scale. But that’s not to say that other forms of genuine sharing don’t continue, like carpooling or couch surfing.

A better recognition of how our employment structures are changing, and how they've moved away from a poetic memory of employment relationships, would give us a better way forward. 

You write about a universal basic income. You call it “automation alimony,” a palliative to deal with a reality where there are fewer jobs. But some people will choose bad work over free money. Do you think we have a cultural adjustment to make to get ready for the post-scarcity economy?
There is this dream in American culture that the individual is someone who can just pull themselves up by their bootstraps and become successful. But I think that people rely on much more than just themselves to become successful. And a better recognition of interconnected societies could help with that.

It’s all great and well to celebrate the individual, but where’s their power? Maybe individualistic politics worked when you could still rely on employers who offered good jobs to lift you up as they became successful. When Sears grew more successful, some of their employees would benefit. But that doesn’t really happen with people who work in an Amazon warehouse.

A better recognition of how our employment structures are changing, and how they’ve moved away from a poetic memory of employment relationships, would give us a better way forward for thinking about how we distribute the benefits of even a jobless future more evenly.

What was the role of technology in the decline of stable work in the last 10 years? Do you feel like it’s fed into the decline? Or has it just responded to it?
I think the opportunities that technology companies have brought in the gig economy have been responsive to declines in traditional employment, actually. And it might be that as unemployment drops, people rely less on gig-work. On the other hand, they might be employed, but employed in jobs that don’t really allow them to pay their bills.

You seem to have a  balanced view of the benefits and risks of these models. How do you weigh the appeal of gig flexibility and rapid on-boarding, against protection for people who don’t have a full time job, don’t have benefits? Is it an inevitable trade off?
I think there’s a bit of a game going on. I’m not sure that being an employee means you can’t have flexibility. Flexibility is not a quality solely in the domain of independent contractor work. Companies have made that argument, that they wouldn’t be able to provide flexibility for employees, but I think that’s junk. All of these questions have the same source, which is, what kind of power asymmetries exist in this world? It’s not that you shouldn’t benefit from flexible, on-demand work. But what power do you have within that system to maintain good working conditions?

It seems like there’s just this tendency of centralization and scale with technology platforms. Do you think it’s a fundamental flaw that Uber is so huge? Is scale itself a problem?
I think monopoly can be a problem. I’m not so sure scale per se is bad. The availability of ride hail services can be a huge boon to people as consumers, and even in the economic opportunity it creates for drivers. The question is always, how you bargain with an algorithm? And that’s not just in the context of Uber. What powers do you have as an individual Facebook user, or user of Google, to bargain? You have none. The question is what kinds of tools we need as a society to say, ‘ok, we want the benefits of platform services and technology. But we also need to ensure a good healthy future, whether it’s for transit, or for democracy.’